Selected material from the Tests of Dyslexia (TOD ) manual. Copyright © 2024 by Western Psychological Services (WPS®).

Provided by
TM

WPS for the sole purpose of introductory reference by qualified professionals. Not to be reprinted, excerpted, or distributed in whole or in
part without the prior written authorization of WPS (rights@wpspublish.com). Full materials available for purchase at wpspublish.com.

5
Psychometric Properties

This chapter presents data to support the reliability and validity of the TOD.
Chapter 4 described the preliminary development of the TOD tests and Rating
Scales and related pilot data, along with the demographic composition of the
standardization and clinical samples. Analyses presented in this chapter were
derived from the standardization and clinical samples described in Chapter 4.

Reliability

Reliability refers to the accuracy or precision of test ability or trait. Internal consistency is estimated as a
scores. Reliability coefficients capture the extent to reliability coefficient, which ranges from 0 to 1. The
which the results are dependable and relatively free methodology for estimating an internal consistency
from error. The standard error of measurement, or coefficient depends upon the type of test. Because
SEM, is derived from statistical estimates of reliabil- of the different types of tests within the TOD, vari-
ity and is frequently used to indicate the precision ous methodologies were used to estimate internal
characterizing an individual score. The smaller the consistency estimates, as described in the following
SEM, the higher the reliability. This section presents paragraphs.
evidence that the TOD test, index, and composite
scores are sufficiently reliable and precise for mea-
Split-Half Reliability
suring an individual’s skills.
Internal consistency reliability for most of the TOD
This first section of the chapter offers a review of
tests was calculated using the split-half method
several reliability concepts and a description of dif-
(Cronbach, 1970). This procedure involves splitting
ferent types of reliability analyses performed for the
test items into halves based on their difficulty. Raw
TOD tests, indexes, composites, and Rating Scales.
scores from the two halves were correlated using the
Pearson product-moment correlation coefficient,
Internal Consistency and then adjusted using the Spearman-Brown for-
mula (Anastasi & Urbina, 1997).
Internal consistency refers to the extent to which all
items in a test or scale consistently measure the same

TOD • W-700M wpspublish.com

 Rasch-Based Reliability adults take the same form. Results illustrate that
all reliability coefficients are ≥.70, with most ≥.80,
Rasch-based reliability was used for two different
indicating acceptable levels of internal consistency
types of tests: tests using item sets and speeded
that support the use of these scores in clinical
(timed) tests. The item set format of Picture Vocabu-
applications.
lary (1S) and Letter and Word Choice (2S) involved
Rasch methodology to derive person ability scores Table 5.3 shows the internal consistency estimates
used to calculate internal consistency. Internal for TOD-C tests in a combined child and adult
consistency reliability for timed tests should not be standardization sample. Internal consistency cannot
calculated using traditional methods because items be calculated for the Oral Reading Efficiency (12C)
are sensitive to both speed and accuracy. Therefore, test because the test functions psychometrically as
the Rasch model was applied to calculate reliability a single “item.” Most reliability coefficients are ≥.90,
for all speeded tests. and almost all are >.80. Less than .02% of analyses
yielded reliabilities lower than .70.
The Rasch analyses yield an estimated ability score
on a logit scale and standard error (SE) for each Tables 5.4 and 5.5 display internal consistency
person in the standardization sample. Each person’s estimates for the Dyslexia Diagnostic Index (DDI),
SE is then squared to produce an error variance Reading and Spelling Index (RSI), and Linguistic
estimate, and the mean error variance (SE2) and the Processing Index (LPI), as well as the composites
variance of the Rasch ability estimate (SD2) are com- that can be calculated using the TOD-S and TOD-C
puted for all persons. Reliability is then estimated as tests, for the child and adult samples, respectively.
r = 1 − (SE2 / SD2). Composite score reliabilities are displayed by grade
for the child sample because TOD-S tests, which are
administered by grade, are included in most of the
Coefficient Alpha
composites. Reliabilities for adults are displayed by
Coefficient alpha (Cronbach, 1988) is the most fre- age range. All composite reliabilities are >.80, and
quently used methodology for estimating reliability most are >.90, indicating excellent reliability.
for rating scales (and other tests without a develop-
Table 5.6 displays the internal consistency estimates
mental gradient) and was applied for the TOD Rating
for the three TOD-C Rating Scales (Self, Parent/Care-
Scales. It represents a more conservative estimate of
giver, and Teacher). All are >.90, indicating excellent
reliability.
reliability.
Tables 5.7 and 5.8 display the internal consistency
Reliability of Linear Combinations
estimates for the TOD-E tests, indexes, and compos-
All indexes and composites were created by sum- ites, and Table 5.9 displays the internal consistency
ming tests, and therefore internal consistency estimates for the two TOD-E Rating Scales (Parent/
reliability for these scores was estimated using Caregiver and Teacher). Almost all internal con-
the formula for reliability of linear combinations sistency estimates are >.90, and all are >.80. These
(Nunnally & Bernstein, 1994). results suggest excellent reliability for the TOD-E
and support the use of these scores in clinical
Tables 5.1 and 5.2 show the internal consistency
applications.
coefficients for the TOD-S tests and Dyslexia Risk
Index (DRI) in the TOD-S child and adult standard- These same reliability analyses were performed
ization samples, respectively. Internal consistency on the TOD clinical samples, and all internal con-
estimates in Table 5.1 are provided by grade because sistency reliabilities for the clinical samples were
the selection of TOD-S form in children up to age 18 consistent with those from the standardization
is determined by grade and not age, whereas those samples.
in Table 5.2 are displayed by age range because all

192 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.1. TOD-S Internal Consistency Estimates and SEMs for Tests and Indexes by Grade: Child

Tests Indexes
PV-S LWC-S WRF-S QRF-S DRI (WRF) DRI (QRF)
Grade n r SEM r SEM r SEM r SEM r SEM r SEM

K 121 .80 6.7 .85 5.8 .93 4.0 — — .93 4.0 — —

1 199 .82 6.4 .88 5.2 .95 3.4 — — .88 5.2 — —

2 221 .74 7.6 .85 5.8 — — .97 2.6 — — .85 5.8

3 170 .72 7.9 .81 6.5 — — .97 2.6 — — .93 4.0
4 146 .78 7.0 .83 6.2 — — .97 2.6 — — .83 6.2
5 140 .83 6.2 .81 6.5 — — .98 2.1 — — .81 6.5

6 145 .72 7.9 .84 6.0 — — .98 2.1 — — .95 3.4

7 128 .74 7.6 .81 6.5 — — .98 2.1 — — .94 3.7
8 103 .72 7.9 .78 7.0 — — .98 2.1 — — .92 4.2
9 101 .70 8.2 .78 7.0 — — .98 2.1 — — .92 4.2
10 88 .81 6.5 .83 6.2 — — .99 1.5 — — .94 3.7
11 87 .80 6.7 .85 5.8 — — .99 1.5 — — .95 3.4
12 74 .83 6.2 .87 5.4 — — .98 2.1 — — .94 3.7

Note. N = 1,723. TOD-S Child internal consistency estimates were calculated by grade because selection of TOD-S form is based on grade (i.e., Grades
K–1, 2–5, and 6–Adult). Internal consistency estimates for all tests were calculated using Rasch-based reliability. Reliability estimates for DRI (WRF)
and DRI (QRF) were calculated using the reliability of linear combinations (Nunnally and Bernstein, 1994). SEM = SD √1 − r, where SEM is the standard
error of measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. PV-S = Picture Vocabulary;
LWC-S = Letter and Word Choice; WRF-S = Word Reading Fluency; QRF-S = Question Reading Fluency; DRI (WRF) = Dyslexia Risk Index (Word
Reading Fluency); DRI (QRF) = Dyslexia Risk Index (Question Reading Fluency).

Table 5.2. TOD-S Internal Consistency Estimates and SEMs for Tests and Index by Age Range: Adult

Tests Index
PV-S LWC-S QRF-S DRI (QRF)
Age (years) n r SEM r SEM r SEM r SEM

18–23 113 .77 7.2 .78 7.0 .99 1.5 .90 4.7
24–39 64 .72 7.9 .70 8.2 .99 1.5 .89 5.0
40–49 40 .74 7.6 .74 7.6 .99 1.5 .85 5.8
50–59 54 .74 7.6 .74 7.6 .99 1.5 .93 4.0
60–69 37 .86 5.6 .84 6.0 .99 1.5 .94 3.7
70–89 39 .88 5.2 .87 5.4 .99 1.5 .93 4.0

Note. N = 347. Internal consistency estimates for all tests were calculated using Rasch-based reliability. Reliability estimate for DRI (QRF)
was calculated using the reliability of linear combinations (Nunnally and Bernstein, 1994). SEM = SD √1 – r, where SEM is the standard
error of measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. PV-S = Picture
Vocabulary; LWC-S = Letter and Word Choice; QRF-S = Question Reading Fluency; DRI (QRF) = Dyslexia Risk Index (Question Reading
Fluency).

Reliability TOD 193

TOD • W-700M wpspublish.com

TOD • W-700M
194 TOD
Table 5.3. TOD-C Internal Consistency Estimates and SEMs for Tests by Age Range: Child and Adult

PHM-C IWS-C RLN-Ca PWR-C WPC-Ca WM-C PAN-C IWR-C BLN-C SEG-C
Age
(years) n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

6 48 .94 3.7 .95 3.4 .97 2.6 .97 2.6 .97 2.6 .75 7.5 .91 4.5 .96 3.1 .65 8.9 .94 3.7
7 81 .96 3.0 .96 3.0 .99 1.5 .97 2.6 .99 1.5 .71 8.1 .85 5.8 .96 3.0 .81 6.5 .88 5.2
8 168 .92 4.2 .96 3.0 .98 2.1 .96 3.0 .99 1.5 .71 8.1 .91 4.5 .90 4.7 .82 6.4 .85 5.8
9 164 .91 4.5 .96 3.0 .98 2.1 .93 4.0 .99 1.5 .70 8.2 .88 5.2 .85 5.8 .87 5.4 .86 5.6
10 132 .89 5.0 .93 4.0 .99 1.5 .94 3.7 .98 2.1 .72 7.9 .82 6.4 .82 6.4 .87 5.4 .88 5.2
11 139 .88 5.2 .90 4.7 .98 2.1 .93 4.0 .99 1.5 .80 6.7 .85 5.8 .88 5.2 .90 4.7 .84 6.0
12 146 .88 5.2 .91 4.5 .99 1.5 .89 5.0 .99 1.5 .75 7.5 .91 4.5 .87 5.4 .86 5.6 .84 6.0
13 121 .82 6.4 .94 3.7 .94 3.7 .91 4.5 .99 1.5 .81 6.5 .85 5.8 .94 3.7 .92 4.2 .93 4.0
14 98 .85 5.8 .87 5.4 .99 1.5 .91 4.5 .99 1.5 .78 7.0 .68 8.5 .92 4.2 .90 4.7 .91 4.5
15 106 .74 7.6 .90 4.7 .99 1.5 .76 7.3 .99 1.5 .78 7.0 .88 5.2 .92 4.2 .86 5.6 .91 4.5
16 82 .74 7.6 .89 5.0 .99 1.5 .87 5.4 .98 2.1 .68 8.5 .88 5.2 .88 5.2 .92 4.2 .91 4.5
17 78 .81 6.5 .93 4.0 .99 1.5 .87 5.4 .96 3.0 .87 5.4 .84 6.0 .94 3.7 .94 3.7 .90 4.7
18 52 .87 5.4 .92 4.2 .99 1.5 .87 5.4 .99 1.5 .87 5.4 .78 7.0 .94 3.7 .87 5.4 .96 3.0
19–23 99 .83 6.2 .91 4.5 .81 6.5 .85 5.8 .98 2.1 .79 6.9 .84 6.0 .94 3.7 .94 3.7 .89 5.0
24–39 64 .77 7.2 .93 4.0 .99 1.5 .84 6.0 .99 1.5 .71 8.1 .82 6.4 .91 4.5 .90 4.7 .87 5.4
40–49 40 .87 5.4 .88 5.2 .99 1.5 .67 8.6 .98 2.1 .83 6.2 .78 7.0 .92 4.2 .82 6.4 .91 4.5
50–59 54 .76 7.3 .93 4.0 .99 1.5 .85 5.8 .98 2.1 .83 6.2 .87 5.4 .91 4.5 .93 4.0 .92 4.2
60–69 37 .93 4.0 .91 4.5 .99 1.5 .91 4.5 .99 1.5 .86 5.6 .85 5.8 .95 3.4 .95 3.4 .94 3.6
70–89 39 .94 3.7 .95 3.4 .99 1.5 .93 4.0 .99 1.5 .83 6.2 .94 3.7 .97 2.6 .97 2.6 .96 3.0

Note. N = 1,748. Internal consistency estimates for timed tests were calculated using Rasch-based reliability; all others were based on the split-half method. SEM = SD √1 − r, where SEM is the standard
error of measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. PHM-C = Phonological Manipulation; IWS-C = Irregular Word Spelling; RLN-C = Rapid
Letter Naming; PWR-C = Pseudoword Reading; WPC-C = Word Pattern Choice; WM-C = Word Memory; PAN-C = Picture Analogies; IWR-C = Irregular Word Reading; BLN-C = Blending; SEG-C = Segmenting;
RWS-C = Regular Word Spelling; SRE1-C = Silent Reading Efficiency Grades 1–5; SRE2-C = Silent Reading Efficiency Grade 6–Adult; RNL-C = Rapid Number and Letter Naming; LM-C = Letter Memory;
RPW-C = Rapid Pseudoword Reading; RIW-C = Rapid Irregular Word Reading; SSL-C = Symbol to Sound Learning; LV-C = Listening Vocabulary; GAN-C = Geometric Analogies.
aTimed test.

Chapter 5 Psychometric Properties

wpspublish.com
TOD • W-700M
Reliability
Table 5.3. TOD-C Internal Consistency Estimates and SEMs for Tests by Age Range: Child and Adult (continued)

RWS-C SRE1/2-Ca RNL-Ca LM-C RPW-Ca RIW-Ca SSL-C LV-C GAN-C

Age
(years) n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

6 48 .91 4.5 .98 2.1 .99 1.5 .71 8.1 .93 4.0 .96 3.0 .95 3.4 .84 6.0 .84 6.0
7 81 .94 3.7 .98 2.1 .99 1.5 .77 7.2 .96 3.0 .99 1.5 .93 4.0 .80 6.7 .93 4.0
8 168 .93 4.0 .98 2.1 .99 1.5 .61 9.4 .95 3.4 .98 2.1 .94 3.7 .82 6.4 .93 4.0
9 164 .96 3.0 .98 2.1 .96 3.0 .80 6.7 .94 3.7 .98 2.1 .95 3.4 .75 7.5 .93 4.0
10 132 .94 3.7 .95 3.4 .99 1.5 .75 7.5 .95 3.4 .98 2.1 .95 3.4 .82 6.4 .89 5.0
11 139 .94 3.7 .89 5.0 .99 1.5 .77 7.2 .96 3.0 .99 1.5 .96 3.0 .82 6.4 .90 4.7
12 146 .94 3.7 .96 3.0 .98 2.1 .80 6.7 .95 3.4 .97 2.6 .95 3.4 .83 6.2 .91 4.5
13 121 .92 4.2 .99 1.5 .97 2.6 .69 8.4 .94 3.7 .98 2.1 .96 3.0 .85 5.8 .89 5.0
14 98 .92 4.2 .98 2.1 .98 2.1 .73 7.8 .96 3.0 .97 2.6 .95 3.4 .88 5.2 .91 4.5
15 106 .87 5.4 .98 2.1 .99 1.5 .70 8.2 .97 2.6 .92 4.2 .95 3.4 .87 5.4 .88 5.2
16 82 .92 4.2 .99 1.5 .99 1.5 .77 7.2 .97 2.6 .95 3.4 .96 3.0 .91 4.5 .89 5.0
17 78 .92 4.2 .99 1.5 .99 1.5 .84 6.0 .96 3.0 .97 2.6 .98 2.1 .91 4.5 .88 5.2
18 52 .94 3.7 .95 3.4 .99 1.5 .81 6.5 .97 2.6 .96 3.0 .97 2.6 .85 5.8 .92 4.2
19–23 99 .90 4.7 .98 2.1 .99 1.5 .77 7.2 .96 3.0 .95 3.4 .94 3.7 .85 5.8 .88 5.2
24–39 64 .82 6.4 .97 2.6 .99 1.5 .84 6.0 .93 4.0 .94 3.7 .97 2.6 .86 5.6 .86 5.6
40–49 40 .88 5.2 .97 2.6 .99 1.5 .84 6.0 .94 3.7 .80 6.7 .96 3.0 .73 7.8 .88 5.2
50–59 54 .92 4.2 .97 2.6 .99 1.5 .84 6.0 .97 2.6 .95 3.4 .96 3.0 .87 5.4 .95 3.4
60–69 37 .91 4.5 .97 2.6 .99 1.5 .89 5.0 .94 3.7 .94 3.7 .99 1.5 .85 5.8 .91 4.5
70–89 39 .94 3.7 .98 2.1 .99 1.5 .58 9.7 .89 5.0 .94 3.7 .97 2.6 .94 3.7 .95 3.4

Note. N = 1,748. Internal consistency estimates for timed tests were calculated using Rasch-based reliability; all others were based on the split-half method. SEM = SD √1 − r, where SEM is the standard
error of measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. PHM-C = Phonological Manipulation; IWS-C = Irregular Word Spelling; RLN-C = Rapid
Letter Naming; PWR-C = Pseudoword Reading; WPC-C = Word Pattern Choice; WM-C = Word Memory; PAN-C = Picture Analogies; IWR-C = Irregular Word Reading; BLN-C = Blending; SEG-C = Segmenting;
RWS-C = Regular Word Spelling; SRE1-C = Silent Reading Efficiency Grades 1–5; SRE2-C = Silent Reading Efficiency Grade 6–Adult; RNL-C = Rapid Number and Letter Naming; LM-C = Letter Memory;
RPW-C = Rapid Pseudoword Reading; RIW-C = Rapid Irregular Word Reading; SSL-C = Symbol to Sound Learning; LV-C = Listening Vocabulary; GAN-C = Geometric Analogies.
aTimed test.

TOD
195

wpspublish.com
TOD • W-700M
196 TOD
Table 5.4. TOD-C Internal Consistency Estimates and SEMs for Indexes and Composites by Grade: Child

Indexes Composites
DDI (WRF) DDI (QRF) LPI RSI (WRF) RSI (QRF) SWA PK BRS DE SP RF (WRF)
Grade n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

1 81 .98 2.1 — — .96 3.0 .98 2.1 — — .98 2.1 .97 2.6 .97 2.6 .98 2.1 .94 3.7 .83 6.2
2 119 — — .97 2.6 .95 3.4 — — .98 2.1 .95 3.4 .98 2.1 .96 3.0 .98 2.1 .96 3.0 — —
3 171 — — .98 2.1 .95 3.4 — — .98 2.1 .97 2.6 .97 2.6 .96 3.0 .98 2.1 .97 2.6 — —
4 147 — — .97 2.6 .95 3.4 — — .98 2.1 .96 3.0 .96 3.0 .94 3.7 .97 2.6 .96 3.0 — —
5 140 — — .97 2.6 .95 3.4 — — .97 2.6 .93 4.0 .98 2.1 .93 4.0 .98 2.1 .96 3.0 — —
6 147 — — .97 2.6 .95 3.4 — — .98 2.1 .88 5.2 .97 2.6 .95 3.4 .98 2.1 .95 3.4 — —
7 131 — — .97 2.6 .94 3.7 — — .97 2.6 .89 5.0 .97 2.6 .95 3.4 .98 2.1 .95 3.4 — —
8 106 — — .97 2.6 .95 3.4 — — .97 2.6 .95 3.4 .97 2.6 .94 3.7 .98 2.1 .94 3.7 — —
9 104 — — .96 3.0 .94 3.7 — — .96 3.0 .87 5.4 .97 2.6 .95 3.4 .98 2.1 .94 3.7 — —
10 91 — — .97 2.6 .96 3.0 — — .97 2.6 .96 3.0 .96 3.0 .95 3.4 .97 2.6 .95 3.4 — —
11 89 — — .98 2.1 .96 3.0 — — .97 2.6 .96 3.0 .97 2.6 .95 3.4 .98 2.1 .96 3.0 — —
12 75 — — .98 2.1 .96 3.0 — — .97 2.6 .97 2.6 .97 2.6 .96 3.0 .98 2.1 .95 3.4 — —

Note. N = 1,401. Internal consistency estimates for composites were calculated using the reliability of linear combinations (Nunnally and Bernstein, 1994). SEM = SD √1 − r, where SEM is the standard
error of measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. DDI (WRF) = Dyslexia Diagnostic Index (Word Reading Fluency); DDI (QRF) = Dyslexia
Diagnostic Index (Question Reading Fluency); LPI = Linguistic Processing Index; RSI (WRF) = Reading and Spelling Index (Word Reading Fluency); RSI (QRF) = Reading and Spelling Index (Question Reading
Fluency); SWA = Sight Word Acquisition composite; PK = Phonics Knowledge composite; BRS = Basic Reading Skills composite; DE = Decoding Efficiency composite; SP = Spelling composite; RF (WRF) =
Reading Fluency composite (Word Reading Fluency); RF (QRF) = Reading Fluency composite (Question Reading Fluency); RC (SRE1) = Reading Comprehension Efficiency 1 composite (Silent Reading
Efficiency Grades 1–5); RC (SRE2) = Reading Comprehension Efficiency 2 composite (Silent Reading Efficiency Grade 6–Adult); PA = Phonological Awareness composite; RAN = Rapid Automatized Naming
composite; AWM = Auditory Working Memory composite; OP = Orthographic Processing composite; VO = Vocabulary composite; RE = Reasoning composite; VR2 = Vocabulary and Reasoning 2 composite;
VR4 = Vocabulary and Reasoning 4 composite.

Chapter 5 Psychometric Properties

wpspublish.com
TOD • W-700M
Reliability
Table 5.4. TOD-C Internal Consistency Estimates and SEMs for Indexes and Composites by Grade: Child (continued)

Composites
RF (QRF) RC (SRE1) RC (SRE2) PA RAN AWM OP VO RE VR2 VR4
Grade n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

1 81 — — — — — — .95 3.4 .99 1.5 .90 4.7 .90 4.7 .86 5.6 .89 5.0 .89 5.0 .91 4.5
2 119 .96 3.0 .98 2.1 — — .95 3.4 .99 1.5 .83 6.2 .81 6.5 .85 5.8 .91 4.5 .91 4.5 .92 4.2
3 171 .96 3.0 .98 2.1 — — .97 2.6 .99 1.5 .88 5.2 .85 5.8 .83 6.2 .92 4.2 .90 4.7 .92 4.2
4 147 .95 3.4 .98 2.1 — — .93 4.0 .99 1.5 .89 5.0 .85 5.8 .86 5.6 .92 4.3 .90 4.7 .92 4.3
5 140 .96 3.0 .98 2.1 — — .95 3.4 .99 1.5 .91 4.5 .84 6.0 .87 5.4 .92 4.3 .91 4.5 .93 4.0
6 147 .96 3.0 — — .99 1.5 .93 4.0 .99 1.5 .89 5.0 .85 5.8 .87 5.4 .93 4.0 .92 4.2 .94 3.7
7 131 .96 3.0 — — .99 1.5 .93 4.0 .99 1.5 .90 4.7 .84 6.0 .86 5.6 .93 4.0 .91 4.5 .94 3.7
8 106 .95 3.4 — — .99 1.5 .94 3.7 .99 1.5 .92 4.2 .87 5.4 .84 6.0 .91 4.5 .92 4.4 .92 4.3
9 104 .95 3.4 — — .99 1.5 .94 3.7 .99 1.5 .89 5.0 .81 6.5 .85 5.8 .91 4.5 .90 4.7 .92 4.3
10 91 .96 3.0 — — .99 1.5 .95 3.4 .99 1.5 .92 4.4 .85 5.8 .87 5.4 .92 4.3 .91 4.5 .92 4.1
11 89 .95 3.4 — — .99 1.5 .96 3.0 .99 1.5 .92 4.3 .85 5.8 .90 4.7 .94 3.7 .92 4.3 .95 3.4
12 75 .96 3.0 — — .99 1.5 .95 3.4 .99 1.5 .93 4.0 .90 4.7 .90 4.7 .93 4.0 .92 4.3 .95 3.4

Note. N = 1,401. Internal consistency estimates for composites were calculated using the reliability of linear combinations (Nunnally and Bernstein, 1994). SEM = SD √1 − r, where SEM is the standard
error of measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. DDI (WRF) = Dyslexia Diagnostic Index (Word Reading Fluency); DDI (QRF) = Dyslexia
Diagnostic Index (Question Reading Fluency); LPI = Linguistic Processing Index; RSI (WRF) = Reading and Spelling Index (Word Reading Fluency); RSI (QRF) = Reading and Spelling Index (Question Reading
Fluency); SWA = Sight Word Acquisition composite; PK = Phonics Knowledge composite; BRS = Basic Reading Skills composite; DE = Decoding Efficiency composite; SP = Spelling composite; RF (WRF) =
Reading Fluency composite (Word Reading Fluency); RF (QRF) = Reading Fluency composite (Question Reading Fluency); RC (SRE1) = Reading Comprehension Efficiency 1 composite (Silent Reading
Efficiency Grades 1–5); RC (SRE2) = Reading Comprehension Efficiency 2 composite (Silent Reading Efficiency Grade 6–Adult); PA = Phonological Awareness composite; RAN = Rapid Automatized Naming
composite; AWM = Auditory Working Memory composite; OP = Orthographic Processing composite; VO = Vocabulary composite; RE = Reasoning composite; VR2 = Vocabulary and Reasoning 2 composite;
VR4 = Vocabulary and Reasoning 4 composite.

TOD 197

wpspublish.com
TOD • W-700M
198 TOD
Table 5.5. TOD-C Internal Consistency Estimates and SEMs for Indexes and Composites by Age Range: Adult

Indexes Composites
DDI (QRF) LPI RSI (QRF) SWA PK BRS DE SP RF (QRF)
Age
(years) n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

18–23 113 .92 4.2 .96 3.0 .97 2.6 .96 3.0 .95 3.4 .95 3.4 .98 2.1 .96 3.0 .96 3.0
24–39 64 .93 4.0 .96 3.0 .97 2.6 .95 3.4 .96 3.0 .93 4.0 .96 3.0 .93 4.0 .96 3.0
40–49 40 .92 4.4 .96 3.0 .95 3.4 .95 3.4 .96 3.0 .91 4.5 .95 3.4 .93 4.0 .97 2.6
50–59 54 .96 3.0 .97 2.6 .98 2.1 .95 3.4 .98 2.1 .93 4.0 .98 2.1 .96 3.0 .96 3.0
60–69 37 .96 3.0 .98 2.1 .98 2.1 .97 2.6 .97 2.6 .97 2.6 .98 2.1 .95 3.4 .96 3.0
70–89 39 .96 3.0 .97 2.6 .98 2.1 .91 4.5 .97 2.6 .97 2.6 .97 2.6 .97 2.6 .96 3.0

Composites
RC (SRE2) PA RAN AWM OP VO RE VR2 VR4
Age
(years) n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

18–23 113 .93 4.0 .94 3.7 .99 1.5 .83 6.2 .93 4.0 .93 4.0 .90 4.7 .93 4.0 .92 4.2
24–39 64 .99 1.5 .95 3.4 .99 1.5 .84 6.0 .91 4.5 .94 3.7 .90 4.7 .93 4.0 .93 4.0
40–49 40 .99 1.5 .95 3.4 .99 1.5 .85 5.8 .92 4.2 .93 4.0 .90 4.7 .94 3.7 .91 4.5
50–59 54 .99 1.5 .97 2.6 .99 1.5 .88 5.2 .93 4.0 .94 3.7 .94 3.7 .95 3.4 .95 3.4
60–69 37 .99 1.5 .98 2.1 .99 1.5 .90 4.7 .94 3.7 .94 3.7 .94 3.7 .95 3.4 .96 3.0
70–89 39 .93 4.0 .98 2.1 .99 1.5 .85 5.8 .92 4.2 .94 3.7 .97 2.6 .94 3.7 .97 2.6

Note. N = 347. Internal consistency estimates for composites were calculated using the reliability of linear combinations (Nunnally and Bernstein, 1994). SEM = SD √1 − r, where SEM is the standard error of
measurement, SD is the standard deviation of the standard score unit (15), and r is the reliability coefficient. DDI (QRF) = Dyslexia Diagnostic Index (Question Reading Fluency); LPI = Linguistic Processing
Index; RSI (QRF) = Reading and Spelling Index (Question Reading Fluency); SWA = Sight Word Acquisition composite; PK = Phonics Knowledge composite; BRS = Basic Reading Skills composite; DE = Decoding
Efficiency composite; SP = Spelling composite; RF (QRF) = Reading Fluency composite (Question Reading Fluency); RC (SRE2) = Reading Comprehension Efficiency 2 composite (Silent Reading Efficiency
Grade 6–Adult); PA = Phonological Awareness composite; RAN = Rapid Automatized Naming composite; AWM = Auditory Working Memory composite; OP = Orthographic Processing composite; VO =
Vocabulary composite; RE = Reasoning composite; VR2 = Vocabulary and Reasoning 2 composite; VR4 = Vocabulary and Reasoning 4 composite.

Chapter 5 Psychometric Properties

wpspublish.com
 Table 5.6. TOD-C Internal Consistency Estimates and SEMs
for the Rating Scale Standardization Sample by Age Range

Self-Rating Parent/Caregiver Rating Teacher Rating

Age (years) n r SEM n r SEM n r SEM

6–7 65 .94 2.4 68 .94 2.4 31 .97 1.7

8–9 210 .91 3.0 186 .94 2.4 93 .96 2.0
10–11 186 .94 2.4 187 .95 2.2 95 .97 1.7
12–13 188 .94 2.4 192 .96 2.0 88 .97 1.7
14–15 147 .94 2.4 130 .93 2.6 42 .95 2.2
16–18 151 .94 2.4 123 .96 2.0 44 .97 1.7
19–23 75 .93 2.6 — — — — — —
24–49 85 .94 2.4 — — — — — —
50–89 94 .95 2.2 — — — — — —

Note. N = 1,452. Parent/Caregiver and Teacher Ratings are for individuals Grades 1–12. Internal consistency estimates
were calculated using Cronbach’s alpha. SEM = SD √1 − r, where SEM is the standard error of measurement, SD is the
standard deviation of the T-score unit (10), and r is the reliability coefficient.

Table 5.7. TOD-E Internal Consistency Estimates and SEMs for Tests by Age

SPW-E RHY-E ERNL-Eb LSW-E ESEG-E LSK-E

Age
(years)a n r SEM r SEM r SEM r SEM r SEM r SEM

5 72 .92 4.2 .96 3.0 .99 1.5 .93 4.0 .96 3.0 .95 3.4
6 122 .93 4.0 .94 3.7 .99 1.5 .97 2.6 .96 3.0 .96 3.0
7 104 .90 4.7 .93 4.0 .97 2.6 .96 3.0 .83 6.2 .91 4.5
8–9:3 44 .87 5.4 .92 4.2 .99 1.5 .92 4.2 .92 4.2 .93 4.0

Note. N = 342. Internal consistency estimates for timed tests were calculated using Rasch-based reliability; all others were based on the split-half
method. SEM = SD √1 − r, where SEM is the standard error of measurement, SD is the standard deviation of the standard score unit (15), and r is the
reliability coefficient. SPW-E = Sounds and Pseudowords; RHY-E = Rhyming; ERNL-E = Early Rapid Number and Letter Naming; LSW-E = Letter and
Sight Word Recognition; ESEG-E = Early Segmenting; LSK-E = Letter and Sound Knowledge.
a8-year normative group extends through age 9 years, 3 months.

bTimed test.

Table 5.8. TOD-E Internal Consistency Estimates and SEMs for Indexes and Composites by Grade

Indexes Composites
EDDI EDDI ERSI ERSI
(WRF) (QRF) ELPI (WRF) (QRF) ESWA EPK EBRS EPA
Grade n r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM r SEM

K 122 .98 2.1 .98 2.1 .98 2.1 .97 2.6 .97 2.6 .93 4.0 .96 3.0 .97 2.6 .96 3.0
1 118 .98 2.1 .98 2.1 .97 2.6 .96 3.0 .96 3.0 .95 3.4 .93 4.0 .97 2.6 .96 3.0
2 102 .97 2.6 .97 2.6 .95 3.4 .96 3.0 .96 3.0 .92 4.2 .93 4.0 .96 3.0 .94 3.7

Note. N = 342. Internal consistency estimates for composites were calculated using the reliability of linear combinations (Nunnally and Bernstein,
1994). SEM = SD √1 − r, where SEM is the standard error of measurement, SD is the standard deviation of the standard score unit (15), and r is the
reliability coefficient. EDDI (WRF) = Early Dyslexia Diagnostic Index (Word Reading Fluency); EDDI (QRF) = Early Dyslexia Diagnostic Index (Question
Reading Fluency); ELPI = Early Linguistic Processing Index; ERSI (WRF) = Early Reading and Spelling Index (Word Reading Fluency); ERSI (QRF) =
Early Reading and Spelling Index (Question Reading Fluency); ESWA = Early Sight Word Acquisition composite; EPK = Early Phonics Knowledge
composite; EBRS = Early Basic Reading Skills composite; EPA = Early Phonological Awareness composite.

Reliability TOD 199

TOD • W-700M wpspublish.com

 Table 5.9. TOD-E Internal Consistency Estimates and SEMs
for the Rating Scale Standardization Sample by Age Range

Parent/Caregiver Rating Teacher Rating

Age (years)a n r SEM n r SEM

5–6 100 .95 2.2 91 .96 2.0

7–9:3 74 .96 2.0 72 .97 1.7

Note. N = 211. Internal consistency estimates were calculated using Cronbach’s alpha. SEM = SD √1 − r,
where SEM is the standard error of measurement, SD is the standard deviation of the T-score unit (10), and r
is the reliability coefficient.
a8-year normative group extends through age 9 years, 3 months.

Test–Retest Reliability standard deviations for the Time 1 and Time 2

standard scores, as well as the effect size of the
Test–retest reliability, also known as temporal
difference between the means. The effect size was
stability, defines the extent to which an examinee’s
calculated as the difference between the mean stan-
test scores remain the same over time, assuming the
dard scores of the two testing occasions, divided by
underlying ability does not change. It is estimated for
the pooled standard deviation. By this method, an
each test by administering the same test to the same
effect size of 0.2 is considered small, 0.5 is considered
individual on two separate occasions, typically only
medium, and 0.8 is considered large (Cohen, 1992).
two or three weeks apart, and then calculating the
The effect sizes range from 0.01 to 0.48 (median =
correlation coefficient between the two sets of scores.
0.18) across the TOD test, index, and composite
Over this brief interval, test scores are not expected
scores. Although most effect sizes are considered
to change appreciably due to development of the
small, indicating negligible change from Time 1 to
underlying abilities. However, scores may change
Time 2 in the average performance of the test–retest
as a result of random variations in performance, or
sample, a few approach the medium range, which
because of learning due to repeated exposure to the
is not unexpected due to the higher likelihood of
same test stimuli.
practice effects with certain tests. Taken as a whole,
A total of 90 individuals participated in test–retest these results support the stability of TOD scores over
studies across the TOD-S, TOD-C, and TOD-E. The time and also indicate the importance of delaying a
TOD-S sample included 81 individuals ranging in second administration for a longer period of time (at
age from 5 to 54 years (M = 14.37 years, SD = 10.93), least three months) to avoid any practice effects.
split evenly between males and females, and was
65% Hispanic and 37% White. In terms of head-of-
Standard Error of Measurement (SEM)
household education level, 52% had a high school
and Confidence Intervals
diploma or lower, and 48% had at least some college.
Sixty-one individuals were in the TOD-C retest sample, The standard error of measurement (SEM) is calcu-
ranging in age from 8 to 54 years (M = 17.43 years, lated from the reliability coefficient and is used to
SD = 11.71), and 30 were in the TOD-E sample, create a confidence interval, i.e., a range of scores
ranging in age from 5 to 8 years (M = 6.38 years, that contains the examinee’s “true score” within a
SD = .92). Because the individuals who took the given probability, e.g., .90. The true score refers to
TOD-C or TOD-E also took the TOD-S, the sample the hypothetical mean that would be obtained from
compositions are demographically similar. repeated testing minus the effects of practice, fatigue,
and other sources of error. The SEM is calculated
Tables 5.10 to 5.12 present the results of the test–
using the formula SEM = SD √1 − r , where SD is the
retest reliability studies for the TOD-S, TOD-C, and
standard deviation of the scale and r is the reliability
TOD-E, respectively. Across all three samples, the
coefficient of the scale. SEM values are displayed
reliability coefficients for tests, indexes, and compos-
in Tables 5.1 to 5.9 next to their respective internal
ites range from .70 to .97, with a median of .88. These
consistency reliability coefficients, and in Tables 5.10
coefficients are satisfactory for tests of developing
to 5.12 next to their respective test–retest reliability
abilities. To illustrate temporal stability in another
coefficients.
way, Tables 5.10 to 5.12 also show the means and

200 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 SEM values can be converted into confidence intervals Rating Scale Cross-Form Consistency
that give a range of probable values for the true
score. For example, the 90% confidence interval rep- Cross-form consistency refers to studies in which
resents the range of scores around the observed score respondents rate an individual on two different
that has a 90% probability of containing the true forms (e.g., parent and teacher ratings). The cross-
score; the 95% confidence interval represents the form ratings may vary because the two respondents
range of scores around the observed score that has are observing the individual being rated in varying
a 95% probability of containing the true score. The environments and at different times. Similarly, an
confidence values for each test are provided in the individual may have a different perspective on them-
scoring tables in the appendices. They are expressed selves than do the adults around them.
in standard score units (T-scores for the Rating Cross-form analyses of the TOD Rating Scales for
Scales) and rounded to the nearest whole number. individuals who had at least two ratings were con-
In most cases, SEMs of internal consistency reliabil- ducted separately for the TOD-C child sample and
ity coefficients are used as the basis for confidence the TOD-E sample. Across all analyses, strong cor-
intervals; however, in the case of speeded tests, test– relations were found between raters: TOD-E Parent/
retest reliability coefficients are used. Rasch-based Caregiver and Teacher Rating forms (r = .75, n = 85);
internal consistency estimates are high (as described TOD-C Parent/Caregiver and Self-Rating forms
previously), and therefore SEMs based on these inter- (r = .63, n = 880); TOD-C Parent/Caregiver and
nal consistency estimates are too narrow to be used Teacher Rating forms (r = .73, n = 344); and TOD-C
in clinical practice. SEMs from Rasch-based internal Self- and Teacher Rating forms (r = .60, n = 391).
consistency reliability estimates for tests with item These results indicate that different raters’ responses
sets are an appropriate basis for confidence values. contribute unique variance, though most scores
Confidence intervals are based on reliability esti- are likely to be similar. In addition, the differences
mates for the whole sample. between ratings by multiple respondents provide
Chapter 2 describes the procedure for using confi- more breadth of information about the individual
dence values to determine confidence intervals, and being rated. Mean T-score differences between raters
Chapter 3 presents interpretation of the confidence were approximately one half of a standard deviation
intervals. for all comparisons (TOD-C Parent/Caregiver and
Teacher Rating difference mean = 4.5; TOD-C Parent/
Caregiver and Self-Rating difference mean = 5.28;
TOD-C Teacher and Self-Rating difference mean = 5.70;
TOD-E Parent/Caregiver and Teacher Rating differ-
ence mean = 4.60).

Table 5.10. TOD-S Test–Retest Reliability: Descriptive Statistics, Effect Sizes, Corrected Correlations, and SEMs

Time 1 Time 2
Effect Corrected
Test/Index Mean SD Mean SD size r ra SEM

Test
Picture Vocabulary 96.85 16.65 99.23 17.94 0.14 .78 .75 7.48
Letter and Word Choice 95.20 15.20 94.35 15.35 0.06 .77 .76 7.29
Word Reading Fluency 98.30 14.09 101.35 14.73 0.22 .93 .94 3.81
Question Reading Fluency 97.10 13.58 98.85 14.70 0.13 .95 .96 3.15

Index
Dyslexia Risk Index (WRF) 96.80 17.47 101.90 16.06 0.29 .95 .94 3.82
Dyslexia Risk Index (QRF) 95.56 13.77 94.98 14.56 0.04 .88 .90 4.80

Note. N = 81. Means, SDs expressed in standard score units (M = 100, SD = 15). Effect size (Cohen’s d) = Time 2 mean minus Time 1 mean, divided by
pooled SD, where pooled SD is √[((Time 1 n) x (Time 1 SD2) + (Time 2 n) x (Time 2 SD2)) / (Time 1 n + Time 2 n)]. WRF = Word Reading Fluency;
QRF = Question Reading Fluency.
aThe reliability coefficient (r) was corrected for variability of normative group (SD = 15) based on standard deviation obtained at Time 1, using

Guilford’s (1954) formula.

Reliability TOD 201

TOD • W-700M wpspublish.com

 Table 5.11. TOD-C Test–Retest Reliability:
Descriptive Statistics, Effect Sizes, Corrected Correlations, and SEMs

Time 1 Time 2
Effect Corrected
Test/Index/Composite Mean SD Mean SD size r ra SEM

Test
Phonological Manipulation 97.85 15.58 99.69 15.54 0.12 .86 .85 5.82
Irregular Word Spelling 96.39 13.46 98.20 14.95 0.13 .89 .91 4.53
Rapid Letter Naming 100.46 14.43 100.31 14.91 0.01 .83 .84 6.00
Pseudoword Reading 99.49 13.98 100.61 12.82 0.08 .71 .74 7.72
Word Pattern Choice 98.92 14.45 105.41 15.35 0.45 .81 .82 6.34
Word Memory 94.38 14.29 96.38 15.47 0.14 .77 .78 7.00
Picture Analogies 93.49 14.02 97.89 13.69 0.31 .72 .74 7.63
Irregular Word Reading 94.31 14.22 97.52 13.63 0.23 .70 .72 7.93
Oral Reading Efficiency 97.61 15.35 101.38 15.01 0.25 .89 .89 5.07
Blending 100.54 15.45 104.20 14.54 0.24 .82 .81 6.47
Segmenting 96.90 17.22 100.03 14.56 0.18 .83 .79 6.94
Regular Word Spelling 97.56 14.18 99.75 14.16 0.15 .93 .94 3.74
Silent Reading Efficiency Grades 1–5 89.80 16.80 95.67 21.72 0.35 .94 .93 4.10
Silent Reading Efficiency Grade 6–Adult 96.98 11.79 101.15 12.45 0.35 .80 .86 5.63
Rapid Number and Letter Naming 100.13 13.12 101.02 15.39 0.07 .80 .84 6.06
Letter Memory 95.26 14.58 97.34 17.12 0.14 .79 .80 6.77
Rapid Pseudoword Reading 99.39 13.77 102.38 15.42 0.22 .79 .81 6.53
Rapid Irregular Word Reading 97.37 10.08 99.20 10.90 0.18 .72 .84 6.08
Symbol to Sound Learning 84.67 14.91 85.79 15.26 0.07 .73 .73 7.82
Listening Vocabulary 94.89 14.26 95.10 13.96 0.02 .88 .89 5.05
Geometric Analogies 95.52 12.96 97.33 15.12 0.14 .85 .88 5.22

Note. N = 61 (Silent Reading Efficiency Grade 6–Adult n = 46). Means, SDs expressed in standard score units (M = 100, SD = 15). Effect size (Cohen’s
d) = Time 2 mean minus Time 1 mean, divided by pooled SD, where pooled SD is √[((Time 1 n) x (Time 1 SD2) + (Time 2 n) x (Time 2 SD2)) / (Time 1 n + Time 2 n)].
Composites including WRF-S and/or SRE1-C [i.e., DDI (WRF), RSI (WRF), RF (WRF), RC (SRE1)] were not included in the analyses due to small sample
sizes. QRF = Question Reading Fluency; SRE2 = Silent Reading Efficiency Grade 6–Adult; WRF-S = Word Reading Fluency; SRE1-C = Silent Reading
Efficiency Grades 1–5; DDI (WRF) = Dyslexia Diagnostic Index (Word Reading Fluency); RSI (WRF) = Reading and Spelling Index (Word Reading
Fluency); RF (WRF) = Reading Fluency composite (Word Reading Fluency); RC (SRE1) = Reading Comprehension Efficiency 1 composite (Silent
Reading Efficiency Grades 1–5).
aThe reliability coefficient (r) was corrected for variability of normative group (SD = 15) based on standard deviation obtained at Time 1, using

Guilford’s (1954) formula.

202 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.11. TOD-C Test–Retest Reliability:
Descriptive Statistics, Effect Sizes, Corrected Correlations, and SEMs (continued)

Time 1 Time 2
Effect Corrected
Test/Index/Composite Mean SD Mean SD size r ra SEM

Index
Dyslexia Diagnostic Index (QRF) 98.13 15.76 95.67 14.93 0.16 .93 .92 4.12
Linguistic Processing Index 100.39 15.84 96.77 15.22 0.23 .89 .88 5.11
Reading and Spelling Index (QRF) 96.13 14.38 95.49 13.65 0.04 .91 .92 4.34

Composite
Sight Word Acquisition composite 98.34 11.65 95.77 10.72 0.22 .81 .87 5.33
Phonics Knowledge composite 101.70 14.83 99.15 14.27 0.17 .79 .80 6.78
Basic Reading Skills composite 98.80 13.46 96.38 12.85 0.18 .76 .79 6.80
Decoding Efficiency composite 101.02 13.49 98.33 12.11 0.20 .79 .82 6.33
Spelling composite 98.93 14.75 96.92 13.77 0.14 .95 .95 3.24
Reading Fluency (QRF) composite 99.75 15.17 95.87 14.53 0.26 .95 .94 3.56
Reading Comprehension Efficiency 2
99.93 13.42 96.26 12.24 0.27 .90 .92 4.24
(SRE2) composite
Phonological Awareness composite 101.25 14.96 97.82 17.09 0.23 .88 .88 5.10
Rapid Automatized Naming composite 101.08 15.35 100.74 13.58 0.02 .84 .84 6.09
Auditory Working Memory composite 96.30 18.01 93.93 15.88 0.13 .84 .79 6.85
Orthographic Processing composite 98.66 15.01 96.79 14.34 0.12 .81 .81 6.58
Vocabulary composite 94.31 15.26 93.49 15.92 0.05 .89 .89 5.07
Reasoning composite 97.35 15.09 93.76 13.87 0.24 .86 .86 5.66
Vocabulary and Reasoning 2 composite 94.92 14.70 92.02 15.79 0.20 .85 .85 5.76
Vocabulary and Reasoning 4 composite 95.40 15.97 92.60 15.41 0.18 .93 .93 4.08

Note. N = 61 (Silent Reading Efficiency Grade 6–Adult n = 46). Means, SDs expressed in standard score units (M = 100, SD = 15). Effect size (Cohen’s
d) = Time 2 mean minus Time 1 mean, divided by pooled SD, where pooled SD is √[((Time 1 n) x (Time 1 SD2) + (Time 2 n) x (Time 2 SD2)) / (Time 1 n + Time 2 n)].
Composites including WRF-S and/or SRE1-C [i.e., DDI (WRF), RSI (WRF), RF (WRF), RC (SRE1)] were not included in the analyses due to small sample
sizes. QRF = Question Reading Fluency; SRE2 = Silent Reading Efficiency Grade 6–Adult; WRF-S = Word Reading Fluency; SRE1-C = Silent Reading
Efficiency Grades 1–5; DDI (WRF) = Dyslexia Diagnostic Index (Word Reading Fluency); RSI (WRF) = Reading and Spelling Index (Word Reading
Fluency); RF (WRF) = Reading Fluency composite (Word Reading Fluency); RC (SRE1) = Reading Comprehension Efficiency 1 composite (Silent
Reading Efficiency Grades 1–5).
aThe reliability coefficient (r) was corrected for variability of normative group (SD = 15) based on standard deviation obtained at Time 1, using

Guilford’s (1954) formula.

Reliability TOD 203

TOD • W-700M wpspublish.com

 Table 5.12. TOD-E Test–Retest Reliability: Descriptive Statistics, Effect Sizes, Corrected Correlations, and SEMs

Time 1 Time 2
Effect Corrected
Test/Index/Composite Mean SD Mean SD size r ra SEM

Test
Sounds and Pseudowords 105.41 14.06 106.97 17.36 0.11 .92 .93 4.03
Rhyming 103.07 14.59 105.55 16.51 0.17 .86 .87 5.41
Early Rapid Number and Letter Naming 103.62 12.64 105.86 12.79 0.18 .82 .86 5.64
Letter and Sight Word Recognition 99.86 12.77 102.66 12.44 0.22 .95 .96 2.94
Early Segmenting 102.20 11.20 107.56 10.58 0.48 .88 .92 4.13
Letter and Sound Knowledge 99.41 15.76 103.93 14.86 0.29 .89 .88 5.12

Index
Early Dyslexia Diagnostic Index (WRF) 96.88 18.42 104.13 17.62 0.39 .98 .97 2.71
Early Linguistic Processing Index 103.60 13.75 108.28 14.11 0.34 .91 .92 4.19
Early Reading and Spelling Index (WRF) 97.32 18.73 104.11 17.33 0.36 .97 .96 3.10

Composite
Early Sight Word Acquisition composite 96.50 16.79 102.57 14.55 0.36 .95 .93 3.83
Early Phonics Knowledge composite 103.07 16.31 106.70 17.88 0.22 .95 .94 3.69
Early Basic Reading Skills composite 100.07 16.27 104.33 15.45 0.26 .94 .92 4.11
Early Phonological Awareness composite 103.20 14.42 108.48 14.09 0.37 .91 .91 4.40

Note. N = 30; some tests/composites have fewer cases. Means, SDs expressed in standard score units (M = 100, SD = 15). Effect size (Cohen’s d) =
Time 2 mean minus Time 1 mean, divided by pooled SD, where pooled SD is √[((Time 1 n) x (Time 1 SD2) + (Time 2 n) x (Time 2 SD2)) / (Time 1 n + Time 2 n)].
Composites including QRF-S [i.e., EDDI (QRF), ERSI (QRF)] were not included in the analyses due to small sample sizes. WRF = Word Reading Fluency;
QRF-S = Question Reading Fluency; EDDI (QRF) = Early Dyslexia Diagnostic Index (Question Reading Fluency); ERSI (QRF) = Early Reading and
Spelling Index (Question Reading Fluency).
aThe reliability coefficient (r) was corrected for variability of normative group (SD = 15) based on standard deviation obtained at Time 1, using

Guilford’s (1954) formula.

204 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Validity

At the most fundamental level, tests and rating scales is especially important given research indicating that
are considered valid if they measure what they are having relatives with reading difficulty is a strong
supposed to measure. Validation evidence must be risk factor for dyslexia (Hamilton & Hayiou-Thomas,
presented for a test’s well-defined purposes, under 2022; Lasnick et al., 2022; Snowling et al., 2019).
specified conditions, and for the populations with
which it is intended to be used. This section presents
Construct Validity
evidence addressing the TOD’s content-description
validity, construct validity, convergent validity, Construct validity is defined as the extent to which
validity based on detection of skill weaknesses, a test (or tests) accurately assesses a theoretical
clinical-groups validity, and predictive validity. construct of interest and is determined by several
sources of evidence. First, because the TOD was
developed to assess reading ability and to be sen-
Content-Description Validity
sitive to the reading limitations associated with
According to Anastasi and Urbina (1997), content- dyslexia, it contains a number of tests that should be
description validity requires “the systematic exami- correlated with reading skill development. Conse-
nation of the test content to determine whether quently, increases in TOD test raw scores should be
it covers a representative sample of the behavior related to chronological age- or grade-level progres-
domain to be measured” (p. 115). Figure 1.1 in sion. TOD test scores should also correlate more
Chapter 1 shows the constructs assessed by the TOD strongly with other measures of reading ability and
and the tests developed to operationalize those related constructs than with theoretically unrelated
constructs. constructs (both within the TOD and when com-
pared with other assessments). Additionally, the
TOD test items were created based on theoretical
factor structure of the TOD should represent the
fit and review of the literature. Test items were
theoretical constructs of dyslexia described in the
constructed to assess the pattern of abilities char-
literature. Finally, TOD scores should differentiate
acterizing dyslexia as described in the research
between examinees known to have reading deficits
literature.
consistent with dyslexia and those who do not.
To ensure content validity of the TOD Rating Scales,
the literature describing characteristics of dyslexia
Developmental Progression
and its underlying etiology was reviewed, along with
other related instruments. Based on these sources, The constructs measured by the TOD display dif-
items were created to elicit the relevant background/ ferent developmental patterns that can provide
history associated with dyslexia and its most salient additional validity support for the tests. All abilities
characterizations (e.g., Kilpatrick, 2015; Mather measured by the TOD should show developmental
& Wendling, in press; Pennington et al., 2019): variability. Thus, prior to creating standard scores,
motivation for reading, general reasoning, verbal TOD raw score means were examined to ensure
comprehension, orthographic processing, phonolog- that they fit with the expectation of skill growth
ical awareness, rapid automatized naming, memory, specific to the construct in question. All of the skills
basic reading skills, reading fluency, reading com- measured by the TOD should show early rapid
prehension, and spelling. Each of the Rating Scales growth that tapers off at different ages. For example,
contains several Yes or No questions related to family although vocabulary knowledge increases through-
history, history of reading support, grade retention, out the life span, it grows quickly beginning at age 3
and previous diagnoses, followed by a set of Likert- and slows down around age 12 for most individuals
type items with responses ranging from Strongly (Byrnes, 2021). Considerable research indicates the
Disagree (1) to Strongly Agree (4); the higher the developmental trajectories that different skills
score on the Rating Scales, the greater the dyslexia should take. The following sections illustrate the ways
risk. Gathering information regarding family history in which the TOD tests conform to expectations.

Validity TOD 205

TOD • W-700M wpspublish.com

 Tests that reach a ceiling during the middle school knowing letter and number names, rapid naming
years are measures of skills that reach mastery and tasks require automaticity. Applying phonological
then do not continue to improve. These skills require skills and knowledge of common English spelling
a shorter growth period to reach proficiency. For patterns is required when reading and spelling both
example, phonological awareness skills follow this pseudowords and irregular words. Reading fluency
pattern. Once children have mastered how to rhyme requires both intact and automatic word recognition.
words, their ability does not continue to improve. These types of tasks have a steeper growth curve
Other phonological skills mastered at younger than early literacy skills, so there is more room to
ages include blending, segmenting, manipulating measure proficiency.
sounds, and knowledge of sound–symbol correspon-
Still other tests continue to show an increase in skills
dences. In a school setting, phonological awareness
beyond high school. These are tests of linguistic
skills and sound–symbol associations are taught in
processing skills (such as letter memory) or acquired
the early elementary grades. Most students will have
knowledge (such as vocabulary). Like those skills
mastered these skills by the middle school years.
that reach an earlier ceiling, these also show rapid
When a particular skill reaches maximum growth as
growth in the early years that slows during high
measured by a particular test and there is no further
school. However, rather than reach a ceiling, the
development of that skill, a ceiling is reached, i.e., the
skills continue to grow until middle age, assuming a
test cannot discriminate skill acquisition any further.
supportive environment. Analysis of these develop-
Most ability tests reflect this phenomenon (e.g., see
mental differences helps explain the ceiling effects
Bracken & McCallum, 2016; McGrew et al., 2014;
evident on some TOD tests.
Wechsler, 2014).
Other tests reach a ceiling during high school years.
Typically, these tests require application of previ-
ously learned skills. For example, in addition to

206 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 TOD-S Figure 5.1 provides an illustration of this increase in
means, using Picture Vocabulary (1S) as an example.
All TOD-S tests demonstrate rapid early growth and
This figure, as well as the others in this section, are
then a slower but consistently increasing trajectory
cut off at age 23 because it represents the plateau of
through high school and beyond. The continued skill
development in early adulthood. The Word Reading
differentiation among the tests demonstrates their
Fluency test (3Sa) covers only three years but shows
utility as a screener across the life span. All TOD-S
consistent growth across the ages of 5–7 years.
test raw scores continue to grow beyond high school.

160

140
Ability Score

120

100

80
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19–23

Age (Years)

Figure 5.1. Increase in Means Example: Picture Vocabulary (Test 1S)

Validity TOD 207

TOD • W-700M wpspublish.com

 TOD-C some skills earlier than others. Table 5.13 identifies
which TOD tests reach a ceiling at middle school,
All TOD-C tests demonstrate rapid early growth and
high school, and beyond. Figures 5.2 to 5.4 illustrate
then a slower but consistently increasing trajectory
examples of these three ceiling categories, showing
through at least middle school. As described above,
the increase in means through age 23, which repre-
it is expected for individuals to obtain mastery in
sents the plateau of development in early adulthood.

Table 5.13. TOD-C Test Ceilings

Middle school High school Beyond high school

Phonological Awareness Auditory Working Memory Spelling

Phonological Manipulation Word Memory Irregular Word Spelling
Blending Letter Memory Regular Word Spelling
Segmenting Rapid Automatized Naming Vocabulary
Visual–Verbal Paired-Associate Learning Rapid Letter Naming Listening Vocabulary
Symbol to Sound Learning Rapid Number and Letter Naming Reasoning
Orthographic Processing Picture Analogies
Word Pattern Choice Geometric Analogies
Phonics Knowledge Word Reading
Pseudoword Reading Irregular Word Reading
Rapid Pseudoword Reading Rapid Irregular Word Reading
Reading Fluency Reading Comprehension Fluency
Oral Reading Efficiency Silent Reading Efficiency

25

20
Raw Score

15

10

5
6 7 8 9 10 11 12 13 14 15 16 17 18 19–23

Age (Years)

Figure 5.2. Ceiling in Middle School Example: Blending (Test 13C)

208 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 120

100

80
Raw Score

60

40

20
6 7 8 9 10 11 12 13 14 15 16 17 18 19–23

Age (Years)

Figure 5.3. Ceiling in High School Example: Rapid Letter Naming (Test 6C)

40

30
Raw Score

20

10

0
6 7 8 9 10 11 12 13 14 15 16 17 18 19–23

Age (Years)

Figure 5.4. Growth Beyond High School Example: Regular Word Spelling (Test 15C)

Validity TOD 209

TOD • W-700M wpspublish.com

 TOD-E followed by a leveling off toward the top of the age
range fits the design and expectations of the tests.
All six TOD-E tests demonstrate a similar progres-
Since the TOD-E is meant to identify struggling read-
sion of means, as illustrated by the Rhyming (5E)
ers, it was important to include second graders even
test examples shown in Figure 5.5. The TOD-E was
though typically developing second graders reached
designed to measure earlier skills displayed in
a ceiling on the TOD-E tests.
kindergarten and first grade by beginning readers;
therefore, a steeper initial progression of scores

25

20
Raw Score

15

10

5
5 6 7 8

Age (Years)

Figure 5.5. TOD-E Ceilings Example: Rhyming (Test 5E)

210 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Test Intercorrelations Table 5.14. Test Intercorrelations: TOD-S Child

The correlations between the individual TOD test PV-S LWC-S WRF-S QRF-S
scores in the standardization sample were examined
PV-S —
to provide further evidence of construct validity.
LWC-S .57 —
Tables 5.14–5.18 display the intercorrelations for
tests in the standardization samples: TOD-S child, WRF-S .39 .55 —
TOD-S adult, TOD-C child, TOD-C adult, and TOD-E. QRF-S .42 .66 NA —
As expected, the tests exhibit correlations that range
considerably, from small (.08) to high (.84). Lower Note. N = 1,723. PV-S = Picture Vocabulary; LWC-S = Letter and Word
Choice; WRF-S = Word Reading Fluency; QRF-S = Question Reading
correlations were found between tests of divergent Fluency.
skill areas. For example, Phonological Manipulation
(4C) and Rapid Number and Letter Naming (17C) Table 5.15. Test Intercorrelations: TOD-S Adult
correlate at .31 in the TOD-C child sample and .30
in the TOD-C adult sample; phonological aware- PV-S LWC-S QRF-S
ness and rapid automatized naming would not be PV-S —
expected to correlate highly. Higher correlations
LWC-S .58 —
were found between tests of similar skills, such as
QRF-S .39 .40 —
Regular Word Spelling (15C) and Irregular Word
Spelling (5C), which correlate at .81 in the TOD-C Note. N = 347. PV-S = Picture Vocabulary; LWC-S = Letter
and Word Choice; QRF-S = Question Reading Fluency.
child sample and .78 in the TOD-C adult sample.
For all tests, the correlations are lower than their
internal consistency reliabilities reported earlier
in the chapter. The correlations are, however, high
enough to warrant their combination to produce
index and composite scores (of combined test scores)
and low enough to show that each test measures a
unique skill and thus can be scored and interpreted
independently.

Validity TOD 211

TOD • W-700M wpspublish.com

 Table 5.16. Test Intercorrelations: TOD-C Child (Including TOD-S)

PV-S LWC-S WRF-S QRF-S PHM-C IWS-C RLN-C PWR-C WPC-C WM-C PAN-C IWR-C

PV-S —
LWC-S .57 —
WRF-S .53 .72 —
QRF-S .39 .53 NA —
PHM-C .28 .33 .59 .36 —
IWS-C .46 .68 .65 .54 .44 —
RLN-C .27 .41 .72 .47 .30 .49 —
PWR-C .36 .48 .74 .48 .47 .62 .52 —
WPC-C .22 .33 .47 .43 .23 .41 .36 .32 —
WM-C .32 .30 .44 .29 .34 .39 .28 .33 .20 —
PAN-C .34 .27 .08 .31 .30 .26 .23 .34 .17 .29 —
IWR-C .49 .63 .81 .53 .39 .68 .49 .61 .31 .31 .32 —
ORE-C .29 .43 .82 .51 .59 .52 .44 .49 .29 .24 .22 .52
BLN-C .29 .25 .28 .24 .34 .34 .25 .33 .16 .28 .30 .33
SEG-C .25 .20 .35 .19 .35 .25 .18 .31 .11 .29 .33 .31
RWS-C .46 .64 .84 .54 .48 .81 .51 .68 .38 .43 .33 .68
SRE1-C .45 .57 .75 .65 .46 .60 .45 .53 .44 .31 .27 .63
SRE2-C .50 .57 NA .69 .47 .60 .44 .51 .44 .35 .38 .55
RNL-C .22 .32 .65 .45 .31 .38 .66 .53 .32 .23 .25 .42
LM-C .26 .32 .42 .29 .30 .43 .27 .33 .24 .62 .23 .33
RPW-C .36 .52 .78 .54 .46 .65 .59 .84 .35 .38 .30 .65
RIW-C .36 .52 .84 .60 .37 .59 .60 .67 .37 .25 .30 .66
SSL-C .27 .25 .33 .26 .34 .30 .24 .40 .18 .37 .27 .29
LV-C .56 .49 .42 .50 .43 .52 .34 .49 .27 .38 .45 .56
GAN-C .41 .36 .53 .36 .38 .39 .27 .43 .16 .42 .49 .37

Note. N = 1,401. PV-S = Picture Vocabulary; LWC-S = Letter and Word Choice; WRF-S = Word Reading Fluency; QRF-S = Question Reading Fluency;
PHM-C = Phonological Manipulation; IWS-C = Irregular Word Spelling; RLN-C = Rapid Letter Naming; PWR-C = Pseudoword Reading; WPC-C = Word
Pattern Choice; WM-C = Word Memory; PAN-C = Picture Analogies; IWR-C = Irregular Word Reading; ORE-C = Oral Reading Efficiency; BLN-C = Blending;
SEG-C = Segmenting; RWS-C = Regular Word Spelling; SRE1-C = Silent Reading Efficiency Grades 1–5; SRE2-C = Silent Reading Efficiency Grade
6–Adult; RNL-C = Rapid Number and Letter Naming; LM-C = Letter Memory; RPW-C = Rapid Pseudoword Reading; RIW-C = Rapid Irregular Word
Reading; SSL-C = Symbol to Sound Learning; LV-C = Listening Vocabulary; GAN-C = Geometric Analogies.

212 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.16. Test Intercorrelations: TOD-C Child (Including TOD-S) (continued)

ORE-C BLN-C SEG-C RWS-C SRE1-C SRE2-C RNL-C LM-C RPW-C RIW-C SSL-C LV-C GAN-C

PV-S
LWC-S
WRF-S
QRF-S
PHM-C
IWS-C
RLN-C
PWR-C
WPC-C
WM-C
PAN-C
IWR-C
ORE-C —
BLN-C .21 —
SEG-C .19 .58 —
RWS-C .50 .39 .34 —
SRE1-C .65 .22 .20 .59 —
SRE2-C .51 .28 .21 .63 NA —
RNL-C .42 .22 .20 .44 .45 .38 —
LM-C .26 .26 .27 .44 .32 .34 .23 —
RPW-C .55 .31 .28 .71 .58 .54 .59 .37 —
RIW-C .55 .24 .20 .60 .63 .58 .62 .24 .72 —
SSL-C .20 .20 .25 .35 .29 .28 .27 .36 .37 .31 —
LV-C .38 .38 .38 .56 .49 .64 .32 .34 .49 .48 .33 —
GAN-C .27 .34 .35 .46 .36 .47 .28 .36 .40 .35 .36 .49 —

Note. N = 1,401. PV-S = Picture Vocabulary; LWC-S = Letter and Word Choice; WRF-S = Word Reading Fluency; QRF-S = Question Reading Fluency;
PHM-C = Phonological Manipulation; IWS-C = Irregular Word Spelling; RLN-C = Rapid Letter Naming; PWR-C = Pseudoword Reading; WPC-C = Word
Pattern Choice; WM-C = Word Memory; PAN-C = Picture Analogies; IWR-C = Irregular Word Reading; ORE-C = Oral Reading Efficiency; BLN-C = Blending;
SEG-C = Segmenting; RWS-C = Regular Word Spelling; SRE1-C = Silent Reading Efficiency Grades 1–5; SRE2-C = Silent Reading Efficiency Grade
6–Adult; RNL-C = Rapid Number and Letter Naming; LM-C = Letter Memory; RPW-C = Rapid Pseudoword Reading; RIW-C = Rapid Irregular Word
Reading; SSL-C = Symbol to Sound Learning; LV-C = Listening Vocabulary; GAN-C = Geometric Analogies.

Validity TOD 213

TOD • W-700M wpspublish.com

 Table 5.17. Test Intercorrelations: TOD-C Adult (Including TOD-S)

PV-S LWC-S QRF-S PHM-C IWS-C RLN-C PWR-C WPC-C WM-C PAN-C IWR-C
PV-S —
LWC-S .58 —
QRF-S .39 .40 —
PHM-C .55 .47 .38 —
IWS-C .56 .66 .53 .50 —
RLN-C .32 .38 .47 .36 .46 —
PWR-C .51 .49 .42 .69 .63 .54 —
WPC-C .33 .39 .43 .40 .36 .29 .38 —
WM-C .49 .43 .34 .49 .49 .40 .45 .32 —
PAN-C .42 .32 .40 .44 .37 .35 .38 .30 .41 —
IWR-C .60 .58 .45 .49 .73 .50 .70 .36 .49 .44 —
ORE-C .38 .42 .51 .43 .56 .50 .43 .32 .36 .33 .52
BLN-C .38 .34 .33 .58 .32 .21 .37 .26 .43 .38 .28
SEG-C .44 .43 .31 .63 .42 .32 .53 .34 .48 .41 .47
RWS-C .60 .63 .52 .60 .78 .52 .71 .44 .53 .43 .72
SRE2-C .58 .50 .69 .51 .60 .48 .56 .47 .47 .45 .60
RNL-C .30 .35 .47 .30 .47 .76 .50 .31 .33 .32 .49
LM-C .40 .41 .38 .46 .46 .37 .47 .29 .65 .41 .48
RPW-C .49 .58 .43 .56 .65 .57 .82 .40 .45 .32 .70
RIW-C .45 .44 .52 .45 .58 .62 .61 .40 .37 .39 .59
SSL-C .50 .45 .31 .54 .42 .29 .47 .27 .53 .41 .47
LV-C .65 .51 .53 .58 .58 .44 .59 .40 .51 .52 .66
GAN-C .53 .37 .36 .55 .46 .40 .51 .32 .53 .66 .51

Note. N = 347. WRF-S and SRE1-C are not taken by adults. PV-S = Picture Vocabulary; LWC-S = Letter and Word Choice; QRF-S = Question Reading
Fluency; PHM-C = Phonological Manipulation; IWS-C = Irregular Word Spelling; RLN-C = Rapid Letter Naming; PWR-C = Pseudoword Reading;
WPC-C = Word Pattern Choice; WM-C = Word Memory; PAN-C = Picture Analogies; IWR-C = Irregular Word Reading; ORE-C = Oral Reading Efficiency;
BLN-C = Blending; SEG-C = Segmenting; RWS-C = Regular Word Spelling; SRE2-C = Silent Reading Efficiency Grade 6–Adult; RNL-C = Rapid Number
and Letter Naming; LM-C = Letter Memory; RPW-C = Rapid Pseudoword Reading; RIW-C = Rapid Irregular Word Reading; SSL-C = Symbol to Sound
Learning; LV-C = Listening Vocabulary; GAN-C = Geometric Analogies; WRF-S = Word Reading Fluency; SRE1-C = Silent Reading Efficiency Grades 1–5.

214 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.17. Test Intercorrelations: TOD-C Adult (Including TOD-S) (continued)

ORE-C BLN-C SEG-C RWS-C SRE2-C RNL-C LM-C RPW-C RIW-C SSL-C LV-C GAN-C
PV-S
LWC-S
QRF-S
PHM-C
IWS-C
RLN-C
PWR-C
WPC-C
WM-C
PAN-C
IWR-C
ORE-C —
BLN-C .29 —
SEG-C .33 .64 —
RWS-C .54 .38 .49 —
SRE2-C .58 .39 .44 .62 —
RNL-C .50 .17 .25 .53 .50 —
LM-C .39 .43 .47 .51 .40 .30 —
RPW-C .50 .31 .47 .71 .55 .56 .44 —
RIW-C .45 .28 .37 .62 .61 .68 .37 .63 —
SSL-C .24 .41 .46 .48 .36 .19 .47 .42 .28 —
LV-C .56 .44 .55 .67 .66 .42 .52 .59 .53 .48 —
GAN-C .32 .42 .51 .49 .44 .30 .45 .42 .43 .53 .56 —

Note. N = 347. WRF-S and SRE1-C are not taken by adults. PV-S = Picture Vocabulary; LWC-S = Letter and Word Choice; QRF-S = Question Reading
Fluency; PHM-C = Phonological Manipulation; IWS-C = Irregular Word Spelling; RLN-C = Rapid Letter Naming; PWR-C = Pseudoword Reading;
WPC-C = Word Pattern Choice; WM-C = Word Memory; PAN-C = Picture Analogies; IWR-C = Irregular Word Reading; ORE-C = Oral Reading Efficiency;
BLN-C = Blending; SEG-C = Segmenting; RWS-C = Regular Word Spelling; SRE2-C = Silent Reading Efficiency Grade 6–Adult; RNL-C = Rapid Number
and Letter Naming; LM-C = Letter Memory; RPW-C = Rapid Pseudoword Reading; RIW-C = Rapid Irregular Word Reading; SSL-C = Symbol to Sound
Learning; LV-C = Listening Vocabulary; GAN-C = Geometric Analogies; WRF-S = Word Reading Fluency; SRE1-C = Silent Reading Efficiency Grades 1–5.

Validity TOD 215

TOD • W-700M wpspublish.com

 Table 5.18. Test Intercorrelations: TOD-E (Including TOD-S)

PV-S LWC-S WRF-S QRF-S SPW-E RHY-E ERNL-E LSW-E ESEG-E LSK-E

PV-S —
LWC-S .58 —
WRF-S .37 .64 —
QRF-S .42 .72 NA —
SPW-E .47 .65 .61 .57 —
RHY-E .42 .52 .39 .49 .61 —
ERNL-E .37 .40 .42 .45 .56 .46 —
LSW-E .47 .66 .68 .70 .76 .57 .64 —
ESEG-E .32 .22 .18 .10 .45 .39 .46 .38 —
LSK-E .45 .55 .51 .62 .72 .60 .59 .73 .52 —

Note. N = 342. PV-S = Picture Vocabulary; LWC-S = Letter and Word Choice; WRF-S = Word Reading Fluency; QRF-S = Question Reading Fluency;
SPW-E = Sounds and Pseudowords; RHY-E = Rhyming; ERNL-E = Early Rapid Number and Letter Naming; LSW-E = Letter and Sight Word Recognition;
ESEG-E = Early Segmenting; LSK-E = Letter and Sound Knowledge.

Confirmatory Factor Analysis (CFA) Evidence the Reading and Spelling Index (RSI) for the TOD-C
Supporting the TOD Diagnostic Indexes and Early Reading and Spelling Index (ERSI) for the
TOD-E. CFA was applied to evaluate the models, with
As described in earlier sections, the TOD tests were
modification if necessary, and analyzed using Mplus
designed to measure the hallmark linguistic risk
(Version 7) software (Muthén & Muthén, 2012). This
factors of dyslexia, specifically limited phonologi-
analytic approach compares the goodness-of-fit sta-
cal awareness, poor orthographic processing, slow
tistics of the one- and two-factor models to evaluate
rapid automatized naming, and limited working
the extent to which these hypothesized models fit the
memory, along with reading and spelling skills
sample data (Byrnes, 2012).
that are typically negatively impacted by dyslexia
(Mather & Wendling, in press; McCallum et al.,
2006). The Dyslexia Diagnostic Index (DDI) and Standardization Sample CFA
Early Dyslexia Diagnostic Index (EDDI) were created
Data defining these factor structures of the TOD-C
based on this research as well as a series of multiple
and TOD-E were taken from the standardization
regression analyses, described in Chapter 4. Because
samples described in Chapter 4. Tables 5.19 and
these diagnostic indicators are the most powerful
5.20 present the model fit statistics, along with factor
(global) predictors within the TOD-C and TOD-E,
loadings and factor correlations, for the TOD-C and
they were subjected to Confirmatory Factor Analysis
TOD-E samples. For both samples, the goodness-of-
(CFA) model testing. That is, two theoretical models
fit statistics represent an acceptable fit of the models
were hypothesized to explain these scores in order to
across the standardization samples. Further, the
evaluate the utility of the overall diagnostic indexes:
one-factor and two-factor models are virtually the
a one-factor model whereby all tests load onto the
same, i.e., the two-factor model does not significantly
overall diagnostic index; and a two-factor model
improve the model fit compared to the one-factor
whereby the tests are separated into the groups
model, and thus interpretation is appropriate using
that make up two component scores, the Linguistic
either the one-factor or two-factor model.
Processing Index (LPI) for the TOD-C and Early Lin-
guistic Processing Index (ELPI) for the TOD-E; and

216 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.19. Comparing Confirmatory Factor Analysis Model Fit for the TOD-C Standardization Sample

One-factor model Two-factor model

Model fit statistics

chi-square 456 452
df 20 19
p <.001 <.001
SRMR .04 .04
RMSEA .11 .11
CFI .92 .92
TLI .89 .88

Dyslexia Diagnostic Index Linguistic Processing Reading and

Factor loadings (DDI) Index (LPI) Spelling Index (RSI)
Letter and Word Choice .72 — .72
Word/Question Reading Fluency .68 — .68
Phonological Manipulation .70 .71 —
Irregular Word Spelling .83 — .84
Rapid Letter Naming .62 .62 —
Pseudoword Reading .75 — .75
Word Pattern Choice .49 .49 —
Word Memory .51 .52 —

Note. N = 1,748. df = degrees of freedom; p = the probability, testing against the null hypothesis, that the RMSEA is zero; SRMR = standardized root-
mean-square residual, average correlation residuals; RMSEA = root-mean-square error of approximation, function of chi-square test of close fit; CFI =
comparative fit index; TLI = Tucker–Lewis index.

Table 5.20. Comparing Confirmatory Factor Analysis Model Fit for the TOD-E Standardization Sample

One-factor model Two-factor model

Model fit statistics

chi-square 149 143
df 20 19
p <.001 <.001
SRMR .05 .05
RMSEA .13 .13
CFI .92 .92
TLI .89 .89

Early Dyslexia Diagnostic Early Linguistic Early Reading and

Factor loadings Index (EDDI) Processing Index (ELPI) Spelling Index (ERSI)
Letter and Word Choice .73 — .73
Word/Question Reading Fluency .70 — .70
Sounds and Pseudowords .86 — .86
Rhyming .68 .71 —
Early Rapid Number and Letter Naming .68 .72 —
Letter and Sight Word Recognition .89 — .89
Early Segmenting .49 .53 —
Letter and Sound Knowledge .83 — .82

Note. N = 342. df = degrees of freedom; p = the probability, testing against the null hypothesis, that the RMSEA is zero; SRMR = standardized root-
mean-square residual, average correlation residuals; RMSEA = root-mean-square error of approximation, function of chi-square test of close fit; CFI =
comparative fit index; TLI = Tucker–Lewis index.

Validity TOD 217

TOD • W-700M wpspublish.com

 Clinical Sample CFA correlations, for the TOD-C and TOD-E clinical
samples. Again, for both samples, the goodness-of-fit
For cross-validation purposes, the same theoretical
statistics represent an acceptable fit of the models,
models were examined in the clinical samples for
and the two-factor model does not significantly
the TOD-C and TOD-E. Due to the lower prevalence
improve the model fit compared to the one-factor
of identified disorders in the adult sample, only the
model. Additionally, the model fit statistics and
child data were included in the clinical data set for
factor loadings are slightly strengthened in the
the TOD-C. Tables 5.21 and 5.22 present the model
clinical sample when compared to the standardiza-
fit statistics, along with factor loadings and factor
tion sample.

Table 5.21. Comparing Confirmatory Factor Analysis Model Fit for the TOD-C Clinical Sample

One-factor model Two-factor model

Model fit statistics

chi-square 229 227
df 20 19
p <.001 <.001
SRMR .04 .04
RMSEA .14 .15
CFI .91 .91
TLI .87 .87

Dyslexia Diagnostic Linguistic Processing Reading and Spelling

Factor loadings Index (DDI) Index (LPI) Index (RSI)
Letter and Word Choice .83 — .83
Word/Question Reading Fluency .83 — .83
Phonological Manipulation .74 .75 —
Irregular Word Spelling .80 — .80
Rapid Letter Naming .79 .80 —
Pseudoword Reading .79 — .79
Word Pattern Choice .61 .62 —
Word Memory .45 .46 —

Note. N = 515. df = degrees of freedom; p = the probability, testing against the null hypothesis, that the RMSEA is zero; SRMR = standardized root-
mean-square residual, average correlation residuals; RMSEA = root-mean-square error of approximation, function of chi-square test of close fit; CFI =
comparative fit index; TLI = Tucker–Lewis index.

218 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.22. Comparing Confirmatory Factor Analysis Model Fit for the TOD-E Clinical Sample

One-factor model Two-factor model

Model fit statistics

chi-square 34 34
df 20 19
p <.028 <.021
SRMR .04 .04
RMSEA .10 .11
CFI .97 .96
TLI .95 .95

Early Dyslexia Early Linguistic Early Reading and

Factor loadings Diagnostic Index (EDDI) Processing Index (ELPI) Spelling Index (ERSI)
Letter and Word Choice .82 — .82
Word/Question Reading Fluency .68 — .68
Sounds and Pseudowords .90 — .89
Rhyming .70 .69 —
Early Rapid Number and Letter Naming .79 .78 —
Letter and Sight Word Recognition .93 — .93
Early Segmenting .66 .65 —
Letter and Sound Knowledge .86 — .87

Note. N = 68. df = degrees of freedom; p = the probability, testing against the null hypothesis, that the RMSEA is zero; SRMR = standardized root-
mean-square residual, average correlation residuals; RMSEA = root-mean-square error of approximation, function of chi-square test of close fit; CFI =
comparative fit index; TLI = Tucker–Lewis index.

Validity TOD 219

TOD • W-700M wpspublish.com

 Convergent Validity are included with the TOD-C and TOD-E analyses.
Due to the large number of TOD tests, only the most
The convergent validation method examines a test’s relevant correlations are presented. These analyses
relationship to other measures of similar constructs. focused on comparing tests of similar skills/con-
It is sometimes referred to as concurrent validity. structs to one another and therefore do not involve
Moderate to strong correlations with convergent the index or composite scores.
measures are seen as supporting the construct valid-
ity of the test under study. This section describes the Each related assessment was taken by a subset of
related assessments that were administered for the individuals from the standardization and clinical
convergent validity study and their correlations with samples. Tables 5.23 (TOD-S/TOD-C) and, later in
the TOD tests that assess similar constructs/skills. this chapter, 5.25 (TOD-S/TOD-E) present the demo-
The TOD-S tests are not presented separately but graphic characteristics for these validation groups.

Table 5.23. Demographic Characteristics of the TOD-S/TOD-C Convergent Validation Samples

WJ IV COG WJ IV ACH CASL-2 CTOPP-2 TOWRE-2 TOC-2 UNIT-GAT

Total sample 48 57 49 48 46 46 42

Gender
Male 22 28 28 16 26 21 14
Female 26 29 21 32 20 25 28

Parents’/Individual’s educational level

No high school diploma 2 3 0 1 0 6 1
High school graduate 12 9 3 8 15 10 15
Some college 13 15 14 10 7 13 3
Bachelor’s degree or higher 21 30 32 29 24 17 23

Race/Ethnicitya
Asian 0 4 2 1 5 2 8
Black/African American 0 2 1 1 2 2 3
White 17 39 34 23 25 21 12
American Indian/Alaska Native 0 0 1 0 7 1 1
Native Hawaiian/Pacific Islander 1 1 0 0 2 0 0
Other/Multiracial 0 0 4 3 0 1 0
Hispanic Origin 30 11 7 20 5 19 18

U.S. geographic region

Northeast 0 5 4 0 0 15 0
Midwest 8 10 30 26 18 6 0
South 31 32 15 14 28 25 42
West 9 10 0 8 0 0 0

Note. WJ IV COG = Woodcock-Johnson IV Tests of Cognitive Abilities; WJ IV ACH = Woodcock-Johnson IV Tests of Achievement; CASL-2 = Comprehensive
Assessment of Spoken Language, Second Edition; CTOPP-2 = Comprehensive Test of Phonological Processing, Second Edition; TOWRE-2 = Test of Word
Reading Efficiency, Second Edition; TOC-2 = Test of Orthographic Competence, Second Edition; UNIT-GAT = Universal Nonverbal Intelligence Test–Group
Abilities Test.
aIndividuals of Hispanic origin are included in the race/ethnicity category under Hispanic Origin; remaining categories include only individuals of non-

Hispanic origin.

220 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.23. Demographic Characteristics of the TOD-S/TOD-C Convergent Validation Samples (continued)

WJ IV COG WJ IV ACH CASL-2 CTOPP-2 TOWRE-2 TOC-2 UNIT-GAT

Age (years)
6 0 0 0 0 1 0 0
7 1 1 0 0 0 0 0
8 3 3 3 2 5 3 5
9 2 4 5 8 1 3 1
10 6 8 10 10 8 5 3
11 3 3 5 2 6 3 2
12 5 10 6 3 5 7 2
13 4 4 4 4 3 4 2
14 2 5 3 2 2 10 3
15 6 5 4 7 1 4 4
16 3 2 3 4 3 1 4
17 2 2 2 6 1 2 4
18 0 0 0 0 3 0 1
19–23 2 2 4 0 7 4 11
24–89 9 8 0 0 0 0 0

Disability status
Clinical 10 23 26 15 12 9 10
Typical 38 34 23 33 34 37 32

Note. WJ IV COG = Woodcock-Johnson IV Tests of Cognitive Abilities; WJ IV ACH = Woodcock-Johnson IV Tests of Achievement; CASL-2 = Comprehensive
Assessment of Spoken Language, Second Edition; CTOPP-2 = Comprehensive Test of Phonological Processing, Second Edition; TOWRE-2 = Test of Word
Reading Efficiency, Second Edition; TOC-2 = Test of Orthographic Competence, Second Edition; UNIT-GAT = Universal Nonverbal Intelligence Test–Group
Abilities Test.
aIndividuals of Hispanic origin are included in the race/ethnicity category under Hispanic Origin; remaining categories include only individuals of non-

Hispanic origin.

Validity TOD 221

TOD • W-700M wpspublish.com

 TOD-C were moderate to strong, providing validation
support for the TOD by the TOC-2 orthographic
For the TOD-C, convergent validity data were
processing tests.
collected from seven assessments: Comprehensive
Assessment of Spoken Language, Second Edition
(CASL-2; Carrow-Woolfolk, 2017); Test of Orthographic WJ IV Tests of Achievement
Competence, Second Edition (TOC-2; Mather et al.,
The WJ IV ACH measures academic achievement
2022); Woodcock-Johnson IV Tests of Achievement
skills. For the purposes of this study, 57 individuals
(WJ IV ACH; Schrank et al., 2014a); Comprehensive
from the TOD-C standardization and clinical sam-
Test of Phonological Processing, Second Edition
ples took five WJ IV ACH tests: Spelling, Letter–Word
(CTOPP-2; Wagner et al., 2013); Woodcock-Johnson
Identification, Word Attack, Passage Comprehension,
IV Tests of Cognitive Abilities (WJ IV COG; Schrank
and Sentence Reading Fluency. The WJ IV ACH Spell-
et al., 2014b); Universal Nonverbal Intelligence
ing test correlated highly with both TOD-C tests of
Test–Group Abilities Test (UNIT-GAT; Bracken &
spelling (Irregular Word Spelling [5C], Regular Word
McCallum, 2019); Test of Word Reading Efficiency,
Spelling [15C]) as well as TOD-S Letter and Word
Second Edition (TOWRE-2; Torgesen et al., 2012).
Choice (2S), a test of spelling recognition. The Letter–
Table 5.24 presents the correlations between the
Word Identification and Word Attack tests correlated
TOD-S and TOD-C tests and their corresponding vali-
at a moderate to high level with the TOD-C tests of
dation tests. Correlations ranged from .30 to .91, and
word reading (Pseudoword Reading [7C], Irregular
most are moderate. As a reminder, a correlation of 0.2
Word Reading [11C]), and the Letter–Word Identi-
is considered small, 0.5 is considered medium/mod-
fication test also correlated highly with Letter and
erate, and 0.8 is considered large/strong (Cohen, 1992).
Word Choice (2S). The WJ IV ACH Passage Com-
prehension test correlated at a moderate level with
CASL-2 both TOD-C reading comprehension tests (Question
Reading Fluency [3Sb], Silent Reading Efficiency
The CASL-2 is an individually administered test of
[16C]). The WJ IV ACH Sentence Reading Fluency
spoken language. The Receptive Vocabulary test
test also correlated at a moderate level with these two
of the CASL-2 was administered to a sample of 49
tests, as well as with TOD-C Oral Reading Efficiency
individuals from the TOD-C standardization and
(12C). These moderate to high correlations between
clinical samples. The Receptive Vocabulary test,
TOD tests and similar constructs on the WJ IV ACH
which requires the examinee to choose which image
provide validation support for the TOD.
matches the word that the examiner says aloud, was
used to provide validation support for the TOD-S
Picture Vocabulary (1S) and the TOD-C Listening CTOPP-2
Vocabulary (22C) tests. Both TOD tests demonstrated
The CTOPP-2 is an individually administered test
a moderate relationship with CASL-2 Receptive
of phonological skills used to determine whether an
Vocabulary.
individual is at risk for reading difficulties. Forty-
eight individuals from the TOD-C standardization
TOC-2 and clinical samples were administered five sub-
tests from the CTOPP-2: Elision, Blending Words,
The TOC-2 assesses orthographic processing skills
Phoneme Isolation, Rapid Digit Naming, and Rapid
that are integral to proficient reading and writing.
Letter Naming. The three CTOPP-2 tests of phono-
Two subtests of the TOC-2, Homophone Spelling and
logical awareness (Elision, Blending Words, and
Letter Choice, were taken by 46 individuals from the
Phoneme Isolation) correlated at a low to moderate
TOD-C standardization and clinical samples. TOC-2
level with the TOD-C tests of phonological awareness
Homophone Spelling, which requires the examinee
(Phonological Manipulation [4C], Blending [13C],
to provide the correct spelling of a homophone
and Segmenting [14C]). The two CTOPP-2 rapid nam-
presented by a picture, was used to validate TOD-S
ing tests correlated moderately with the two TOD-C
Letter and Word Choice (2S). TOC-2 Letter Choice is
rapid naming tests (Rapid Letter Naming [6C], Rapid
a timed test in which the examinee completes words
Number and Letter Naming [17C]).
with a missing letter (b, d, p, or q) and was compared
with TOD-S Letter and Word Choice (2S), as well as
TOD-C Word Pattern Choice (8C). All correlations

222 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.24. Convergent Validation Correlations: TOD-S/TOD-C

TOD • W-700M
Validity
CASL-2 TOC-2 WJ IV ACH
Sentence
Receptive Homophone Letter–Word Passage Reading
TOD-S/TOD-C Testa Vocabulary Spelling Letter Choice Spelling ID Word Attack Comprehension Fluency

Picture Vocabulary (1S) .49

Letter and Word Choice (2S) .74 .56 .81 .65
Question Reading Fluency (3Sb) .43 .54
Phonological Manipulation (4C)
Irregular Word Spelling (5C) .90
Rapid Letter Naming (6C)
Pseudoword Reading (7C) .65 .73
Word Pattern Choice (8C) .46
Word Memory (9C)
Picture Analogies (10C)
Irregular Word Reading (11C) .76 .72
Oral Reading Efficiency (12C) .45
Blending (13C)
Segmenting (14C)
Regular Word Spelling (15C) .91
Silent Reading Efficiency Grade .52 .69
6–Adult (16C)
Rapid Number and Letter Naming (17C)
Letter Memory (18C)
Rapid Pseudoword Reading (19C)
Rapid Irregular Word Reading (20C)
Symbol to Sound Learning (21C)
Listening Vocabulary (22C) .51
Geometric Analogies (23C)

Note. n varies by test: CASL-2 n = 49; TOC-2 n = 46; WJ IV ACH n = 57; CTOPP-2 n = 48; WJ IV COG n = 48; UNIT-GAT n = 42; TOWRE-2 n = 46. All correlations significant at <.01 except CTOPP-2 Elision,
Blending Words, and Phoneme Isolation with TOD-C Blending and Segmenting significant at >.05. CASL-2 = Comprehensive Assessment of Spoken Language, Second Edition; TOC-2 = Test of Orthographic
Competence, Second Edition; WJ IV ACH = Woodcock-Johnson IV Tests of Achievement; CTOPP-2 = Comprehensive Test of Phonological Processing, Second Edition; WJ IV COG = Woodcock-Johnson IV
Tests of Cognitive Abilities; UNIT-GAT = Universal Nonverbal Intelligence Test–Group Abilities Test; TOWRE-2 = Test of Word Reading Efficiency, Second Edition.
aSample sizes for Word Reading Fluency and Silent Reading Efficiency Grades 1–5 were too small for correlational analyses.

Table 5.24 continued on next page

TOD 223

wpspublish.com
 Table 5.24. Convergent Validation Correlations: TOD-S/TOD-C (continued)

TOD • W-700M
CTOPP-2 WJ IV COG UNIT-GAT TOWRE-2

224 TOD
Visual Object
Blending Phoneme Rapid Digit Rapid Letter Numbers Auditory Number
TOD-S/TOD-C Testa Elision Words Isolation Naming Naming Reversed Learning Sequencing Full Scale Full Scale

Picture Vocabulary (1S)

Letter and Word Choice (2S)
Question Reading Fluency (3Sb)
Phonological Manipulation (4C) .61 .45 .44
Irregular Word Spelling (5C)
Rapid Letter Naming (6C) .54 .42
Pseudoword Reading (7C)
Word Pattern Choice (8C)
Word Memory (9C) .57 .53
Picture Analogies (10C) .47
Irregular Word Reading (11C)
Oral Reading Efficiency (12C) .55
Blending (13C) .30 .50 .38
Segmenting (14C) .36 .35 .34
Regular Word Spelling (15C)
Silent Reading Efficiency Grade
6–Adult (16C)
Rapid Number and Letter Naming (17C) .56 .59
Letter Memory (18C) .57 .46
Rapid Pseudoword Reading (19C) .82
Rapid Irregular Word Reading (20C) .82
Symbol to Sound Learning (21C) .56
Listening Vocabulary (22C)
Geometric Analogies (23C) .58

Note. n varies by test: CASL-2 n = 49; TOC-2 n = 46; WJ IV ACH n = 57; CTOPP-2 n = 48; WJ IV COG n = 48; UNIT-GAT n = 42; TOWRE-2 n = 46. All correlations significant at <.01 except CTOPP-2 Elision,
Blending Words, and Phoneme Isolation with TOD-C Blending and Segmenting significant at >.05. CASL-2 = Comprehensive Assessment of Spoken Language, Second Edition; TOC-2 = Test of Orthographic
Competence, Second Edition; WJ IV ACH = Woodcock-Johnson IV Tests of Achievement; CTOPP-2 = Comprehensive Test of Phonological Processing, Second Edition; WJ IV COG = Woodcock-Johnson IV
Tests of Cognitive Abilities; UNIT-GAT = Universal Nonverbal Intelligence Test–Group Abilities Test; TOWRE-2 = Test of Word Reading Efficiency, Second Edition.
aSample sizes for Word Reading Fluency and Silent Reading Efficiency Grades 1–5 were too small for correlational analyses.

Chapter 5 Psychometric Properties

wpspublish.com
 WJ IV Tests of Cognitive Abilities TOD-E
Three subtests from the WJ IV COG were taken by For the TOD-E, convergent validity data were collected
48 individuals from the TOD-C standardization for three of the same assessments that were used
and clinical samples. Two of these tests, Numbers for the TOD-C convergent validity study: Compre-
Reversed and Object Number Sequencing, mea- hensive Assessment of Spoken Language, Second
sure auditory working memory and demonstrated Edition (CASL-2; Carrow-Woolfolk, 2017); Woodcock-
moderate correlations with the two TOD-C tests of Johnson IV Tests of Achievement (WJ IV ACH;
working memory, Word Memory (9C) and Letter Schrank et al., 2014a); and Comprehensive Test of
Memory (18C). The Visual Auditory Learning test on Phonological Processing, Second Edition (CTOPP-2;
the WJ IV COG correlated moderately with Symbol Wagner et al., 2013). Table 5.25 presents the demo-
to Sound Learning (21C), both of which are tests of graphic characteristics for the TOD-E validation
visual–verbal paired-associate learning. group. Table 5.26 presents the correlations between
the TOD-S and TOD-E tests and their corresponding
validation tests.
UNIT-GAT
The UNIT-GAT is a nonverbal screener of reasoning
CASL-2
with two subtests, Analogic Reasoning and Quan-
titative Reasoning. Forty-two individuals from the The CASL-2 Receptive Vocabulary test was adminis-
TOD-C standardization and clinical samples took tered to a sample of 33 individuals from the TOD-E
the UNIT-GAT. The UNIT-GAT is intended to be a standardization and clinical samples and correlated
screener, and interpretation at the full-scale level moderately with TOD-S Picture Vocabulary (1S).
is most relevant, rather than consideration of the
relationship among the individual subtests and the
WJ IV ACH
TOD-C scores. Consequently, the correlation between
the Full Scale score and the two TOD-C reasoning Three of the tests from the WJ IV ACH used to
tests (Picture Analogies [10C], Geometric Analogies validate the TOD-C tests were also taken by 50 indi-
[23C]) were of primary interest; moderate correla- viduals in the TOD-E standardization and clinical
tions were obtained. samples: Letter–Word Identification, Spelling, and
Word Attack. As in the TOD-C sample study, both
Letter–Word Identification and Spelling correlated
TOWRE-2
highly with Letter and Word Choice (2S). Letter–Word
The TOWRE-2 was taken by 46 individuals in the Identification also correlated highly with Letter and
TOD-C standardization and clinical samples. It Sight Word Recognition (7E), and Word Attack cor-
assesses reading efficiency in two subtests: Sight related highly with Sounds and Pseudowords (4E).
Word Efficiency, which requires the examinee to
read real words in 45 seconds; and Phonemic Decod-
CTOPP-2
ing Efficiency, which requires reading nonwords in
45 seconds. These two subtests combine into a single Four of the subtests from the CTOPP-2 used to
full-scale score. The TOWRE-2 full-scale score had validate the TOD-C tests were also taken by 31 indi-
high correlations with the TOD-C rapid word read- viduals from the TOD-E standardization and clinical
ing tests (Rapid Pseudoword Reading [19C], Rapid samples: Elision, Blending Words, Rapid Digit Naming,
Irregular Word Reading [20C]) and a moderate cor- and Rapid Letter Naming. Elision correlated mod-
relation with the Oral Reading Efficiency (12C) test. erately with Rhyming (5E), while Blending Words
demonstrated moderate to high correlations with
Early Segmenting (8E) and Letter and Sound Knowl-
edge (9E). The two CTOPP-2 rapid naming tests
correlated moderately with the TOD-E Early Rapid
Number and Letter Naming (6E) test.

Validity TOD 225

TOD • W-700M wpspublish.com

 Table 5.25. Demographic Characteristics of the TOD-S/TOD-E Convergent Validation Samples

WJ IV ACH CASL-2 CTOPP-2

Total sample 50 33 31

Gender
Male 22 12 17
Female 28 21 14

Parents’ educational level

No high school diploma 6 0 1
High school graduate 9 0 6
Some college 12 10 11
Bachelor’s degree or higher 23 23 13

Race/Ethnicitya
Asian 8 8 4
Black/African American 9 0 2
White 11 16 11
Other/Multiracial 3 1 2
Hispanic Origin 19 8 12

U.S. geographic region

Northeast 26 4 0
Midwest 15 9 18
South 8 13 5
West 1 7 8

Age (years)b
5 14 3 11
6 12 7 8
7 16 15 10
8–9:3 8 8 2

Disability status
Clinical 22 7 8
Typical 28 26 23

Note. WJ IV ACH = Woodcock-Johnson IV Tests of Achievement; CASL-2 = Comprehensive Assessment of Spoken Language, Second Edition; CTOPP-2 =
Comprehensive Test of Phonological Processing, Second Edition.
aIndividuals of Hispanic origin are included in the race/ethnicity category under Hispanic Origin; remaining categories include only individuals of non-

Hispanic origin.
b8-year validation group extends through age 9 years, 3 months.

226 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

TOD • W-700M
Validity
Table 5.26. Convergent Validation Correlations: TOD-S/TOD-E

CASL-2 WJ IV ACH CTOPP-2

Receptive Letter–Word Blending Rapid Digit Rapid Letter
TOD-S/TOD-E Testa Vocabulary Identification Spelling Word Attack Elision Words Naming Naming

Picture Vocabulary (1S) .52

Letter and Word Choice (2S) .86 .85
Sounds and Pseudowords (4E) .90
Rhyming (5E) .62
Early Rapid Number and Letter .76 .71
Naming (6E)
Letter and Sight Word Recognition (7E) .91
Early Segmenting (8E) .67
Letter and Sound Knowledge (9E) .80

Note. n varies by test: CASL-2 n = 33; WJ IV ACH n = 50; CTOPP-2 n = 31. CASL-2 = Comprehensive Assessment of Spoken Language, Second Edition; WJ IV ACH = Woodcock-Johnson IV Tests of
Achievement; CTOPP-2 = Comprehensive Test of Phonological Processing, Second Edition.
aSample size for Word Reading Fluency and Question Reading Fluency in the TOD-E sample was too small to conduct validation analysis; evidence in the TOD-C sample supports their validity.

TOD 227

wpspublish.com
 Detection of Skill Weaknesses results for multiple values so that clinicians can
choose a cutoff score that is best suited to their
The TOD was designed to detect weaknesses in
clinical population.
abilities and skills associated with dyslexia and to
aid examiners in screening, diagnosing, and plan- To illustrate, using a cutoff of 80 (one and a third
ning interventions. In particular, the TOD-S Dyslexia standard deviations below the mean) for the TOD-S
Risk Index (DRI), TOD-C Dyslexia Diagnostic Index DRI yields a sensitivity value of .80 and specificity
(DDI), and TOD-E Early Dyslexia Diagnostic Index of .99. In practical terms, this means that 80% of the
(EDDI) were created to differentiate between individ- individuals with clinical diagnoses associated with
uals either having or being at risk for having dyslexia dyslexia had standard scores less than or equal to
and those with typical reading skills. 80, whereas 99% of the typically developing children
had standard scores greater than 80. Using a very
Conditional probability analyses (also known as
strict guideline for eligibility, such as a standard
receiver operating characteristic [ROC] curves) were
score of 70 or less, the specificity is also .99 (i.e., only
run to determine the capacity of the TOD to detect
1% or fewer of typically developing children had
skill deficits associated with dyslexia at various
standard scores of 70 or less, which is ≥2 SD below
cutoff values. For these analyses, children diagnosed
the mean). However, due to the variability inher-
with a learning disability in reading were compared
ent in clinical data, only the most severely impaired
to the standardization sample of typically develop-
individuals will be identified as having dyslexia
ing children. Analyses were obtained from the DRI
when using such a strict cutoff value (e.g., sensitiv-
for the TOD-S, the DDI for the TOD-C, and the EDDI
ity of .40 for the TOD-S and the TOD-C). This finding
for the TOD-E.
demonstrates that a cutoff score of 80 provides a
Results indicated that each measure of the TOD reasonable balance between identifying individuals
risk and diagnostic scores provided statistically with dyslexia, while not overidentifying those indi-
significant improvement over chance in detecting viduals who do not have dyslexia.
dyslexia status: TOD-S DRI score (area under ROC
These results serve as a reminder that at any level of
curve = .972, p < .001); TOD-C DDI score (area under
test score interpretation, there is a risk of under- or
ROC curve = .989, p < .001); TOD-E EDDI score (area
overidentifying children who are in need of inter-
under ROC curve = .989, p < .001).
vention. Although the TOD provides a measurement
Tables 5.27 to 5.29 display the sensitivity and speci- of skill difficulties associated with dyslexia, results
ficity associated with various standard score (SS) should not be used in isolation for diagnosis or treat-
values of the TOD. Sensitivity refers to a test’s capacity ment planning. Instead, these results should be used
to detect true positive cases of the condition in ques- in concert with other data (e.g., TOD Rating Scales,
tion, i.e., dyslexia. Specificity refers to a test’s capacity parent and teacher interview, review of available
to exclude true negative cases (persons who do not records, direct observation, and other assessment
have the condition in question). Betz et al. (2013) results, if available).
recommend providing sensitivity and specificity

228 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.27. Conditional Probability Analysis for
Detection of Clinical Cases Using the TOD-S
Dyslexia Risk Index (DRI) Standard Score

SS cutoff Sensitivity Specificity

70 .40 .99
75 .58 .99
80 .80 .99
85 .93 .96
90 .99 .87

Note. The analyzed sample included 179 clinically diagnosed children and
1,486 typically developing children.

Table 5.28. Conditional Probability Analysis for

Detection of Clinical Cases Using the TOD-C
Dyslexia Diagnostic Index (DDI) Standard Score

SS cutoff Sensitivity Specificity

70 .40 .99
75 .54 .99
80 .78 .97
85 .94 .91
90 .99 .82

Note. The analyzed sample included 160 clinically diagnosed children and
1,285 typically developing children.

Table 5.29. Conditional Probability Analysis for

Detection of Clinical Cases Using the TOD-E
Early Dyslexia Diagnostic Index (EDDI) Standard Score

SS cutoff Sensitivity Specificity

70 .34 .99
75 .63 .99
80 .80 .99
85 .98 .94
90 .99 .84

Note. The analyzed sample included 21 clinically diagnosed children and

249 typically developing children.

Validity TOD 229

TOD • W-700M wpspublish.com

 Validity Evidence for Clinical Groups and effect sizes for the comparisons between this
clinical group and their corresponding matched
An important practical aspect of validity is the control group. The expectation was that the mea-
capacity of test scores to distinguish typically devel- sures of reading and spelling would show large effect
oping individuals from individuals who are expected sizes of the differences between group means, while
to perform differently in the measured ability. the linguistic processing, vocabulary, and reasoning
In analyzing the clinical groups for the TOD, a ran- measures would have smaller differences in effect
domized, matched control group was drawn from sizes. The results support this expectation.
the typical sample, separately for each comparison The effect sizes of the differences between group
of group means. Each clinical case was paired with a means for the reading and spelling tests were all
case of the same age, gender, and parent education large, ranging from 1.05 to 1.50, while those for the
level. The means of the two groups were then com- linguistic processing, vocabulary, and reasoning
pared across TOD test, index, and composite scores. tests were medium to large, ranging from 0.31 to
Effect sizes are reported to determine whether the 1.11. The effect sizes for the DRI and DDI were also
group differences are large enough to be considered both large, 1.42 and 1.41, respectively, reflecting the
clinically meaningful. As previously noted, an effect validity of the risk and diagnostic scores to differen-
size of 0.2 is considered small, 0.5 is considered tiate between individuals diagnosed with dyslexia or
medium, and 0.8 is considered large (Cohen, 1992). a learning disability in reading and those who were
By convention, an effect size is considered clinically not. Similarly, effect sizes of the mean differences in
meaningful only if it is medium or larger in magni- the index and composite scores measuring reading
tude. In this analysis, scores from TOD-S are reported and spelling skills were large, ranging from 1.31 to
along with the TOD-C and TOD-E samples. 1.71, while those for index and composite scores
measuring other skills were medium to large, rang-
TOD-C Child Clinical Sample ing from 0.57 to 1.12.

The TOD-C child clinical sample included 511 These results provide further validation for the
individuals ages 6–18 years. Chapter 4 describes the TOD-C by illustrating that the biggest differences in
sample demographics and diagnostic breakdown. scores between individuals with dyslexia or a learn-
For the clinical discrimination study, this sample was ing disability in reading and their matched controls
divided into seven groups. Five of these groups were were in reading and spelling; these are the precise
expected to demonstrate differences in TOD scores skills in which individuals with these diagnoses have
when compared with a matched typically developing the greatest difficulty. Overall, the TOD-C scores dis-
sample: reading learning disability (RLD), language tinguish well between individuals who are at risk for
disorder, attention-deficit/hyperactivity disorder having dyslexia or a learning disability in reading
(ADHD), autism spectrum disorder (ASD), and intel- and those who are not.
lectual disability (ID) and developmental delay (DD).
Two of the groups were expected to show minimal Language Disorder
differences in TOD scores: speech disorder and a
combined group that included emotional disorders, Another clinical group of interest for the TOD is
deaf/hard of hearing, visually impaired, and other individuals diagnosed with a language disorder.
health/mental health conditions not accounted for Table 5.31 shows the descriptive statistics and effect
by any other group. sizes for the comparisons between this clinical group
of 33 individuals and their corresponding matched
control group. Because individuals with a develop-
Reading Learning Disability mental language disorder are heterogenous in terms
The primary clinical group of interest for the TOD of the manifestation of the disorder, some may have
consists of 278 individuals diagnosed with dyslexia specific difficulty with reading and spelling, whereas
or a learning disability in reading. (Note that some others may not. Thus, the expectation was that the
individuals from this group had comorbid clinical tests of reading and spelling would show larger effect
diagnoses and thus are represented in more than sizes between group means than the linguistic pro-
one group.) Table 5.30 shows the descriptive statistics cessing, vocabulary, and reasoning tests. However,

230 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 the differences were not expected to be as large in of the mean differences in the index and composite
magnitude as for the RLD sample. scores measuring reading and spelling skills were
medium to large, ranging from 0.61 to 0.90. Effect
The effect sizes of the differences between group
sizes of mean differences of index and composite
means for the reading and spelling tests were
scores measuring other skills were also medium to
medium to large, ranging from 0.62 to 1.17, while
large though a bit smaller in magnitude, ranging
those for the linguistic processing, vocabulary, and
from 0.42 to 0.81.
reasoning tests were small to large, ranging from
0.20 to 1.00. The effect sizes for the DRI and DDI Overall, these effect sizes demonstrate the ability of
were also both large, 0.70 and 0.93, respectively. the TOD tests to distinguish individuals with ADHD
Similarly, effect sizes of the mean differences in the who have weaknesses in reading and spelling and
index and composite scores measuring reading and linguistic processing from those who do not.
spelling skills were large, ranging from 0.78 to 1.01,
while those for index and composite scores measur-
Autism Spectrum Disorder
ing other skills were medium to large, ranging from
0.49 to 0.84. Individuals who are diagnosed with autism spectrum
disorder (ASD) often show difficulties in language
These results support the ability of the TOD tests,
and related skills. Table 5.33 shows the descriptive
indexes, and composites in the skill areas of reading
statistics and effect sizes for the 49 individuals diag-
and spelling to distinguish individuals who have
nosed with ASD and their corresponding matched
developmental language disorders from those who
sample. The results show a similar range of effect
do not. They also distinguish well between the two
sizes across all tests, indexes, and composites. The
groups in most other associated skills of linguistic
effect sizes of the mean differences between tests
processing, vocabulary, and reasoning.
ranged from small to large for the reading and
spelling tests (0.31 to 1.02), as well as for the tests
ADHD of linguistic processing, vocabulary, and reasoning
(0.33 to 0.87). Effect sizes for the index and compos-
It is not unusual for individuals with ADHD to have
ite scores were mostly in the medium range, though
difficulties in reading and spelling due to high
a few were large (0.45 to 0.84).
comorbidity between these two disorders. However,
challenges for individuals with ADHD often extend These results indicate that the TOD is sensitive to
to other skill areas. Such individuals also often the difficulties that are often present for individuals
have areas of strength in which they perform quite with ASD.
similarly to their typically developing peers. Because
ADHD does not necessarily affect one specific skill,
Intellectual Disability and Developmental Delay
the expectation was to find a range of effect sizes
when comparing 118 individuals with a primary Individuals with a diagnosis of intellectual disability
diagnosis of ADHD with a matched control group. (ID) or developmental delay (DD) generally demon-
Table 5.32 shows the descriptive statistics and effect strate skill deficits across most or all skills measured
sizes for these comparisons. by the TOD. Thirty-four individuals with a diagnosis
of ID or DD were compared with a matched control
The effect sizes of the differences between group
group. Table 5.34 shows the descriptive statistics and
means for the reading and spelling tests were
effect sizes for these comparisons. Effect sizes for all
medium to large, ranging from 0.53 to 0.90, while
tests, indexes, and composites were large, ranging
those for the linguistic processing, vocabulary, and
from 0.69 to 1.98. This indicates that the TOD test,
reasoning tests were small to large, ranging from 0.00
index, and composite scores distinguish meaning-
to 0.79. The effect sizes for the DRI and DDI were
fully between typically developing individuals and
both large, 0.79 and 1.00, respectively. Effect sizes
those diagnosed with ID or DD.

Validity TOD 231

TOD • W-700M wpspublish.com

 Table 5.30. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With a Reading Learning Disability (RLD) and Matched Control Group

Matched control
Individuals with RLD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Test
Picture Vocabulary 268 86.22 20.10 102.36 14.03 0.80
Letter and Word Choice 268 82.81 15.25 101.31 14.04 1.21
Question Reading Fluency 256 84.51 16.71 102.58 13.69 1.08
Phonological Manipulation 276 85.58 16.59 102.94 14.65 1.05
Irregular Word Spelling 277 82.60 15.79 102.05 14.50 1.23
Rapid Letter Naming 274 82.84 16.72 101.41 14.26 1.11
Pseudoword Reading 276 83.19 15.09 102.68 13.49 1.29
Word Pattern Choice 277 92.17 16.33 101.08 15.11 0.55
Word Memory 276 93.39 14.73 101.34 13.94 0.54
Picture Analogies 277 90.60 19.82 102.12 14.16 0.58
Irregular Word Reading 277 80.52 17.65 100.75 14.08 1.15
Oral Reading Efficiency 276 83.84 16.65 101.95 14.18 1.09
Blending 277 91.96 22.32 101.37 15.42 0.42
Segmenting 277 93.85 22.36 100.83 14.95 0.31
Regular Word Spelling 276 83.10 14.80 102.35 14.57 1.30
Silent Reading Efficiency Grades 1–5 155 80.72 19.32 102.24 14.94 1.11
Silent Reading Efficiency Grade 6–Adult 116 85.80 15.84 102.46 13.38 1.05
Rapid Number and Letter Naming 277 84.15 17.24 102.73 14.20 1.08
Letter Memory 277 93.12 15.56 100.77 13.19 0.49
Rapid Pseudoword Reading 246 83.54 12.88 102.86 14.25 1.50
Rapid Irregular Word Reading 260 82.56 14.20 101.70 13.01 1.35
Symbol to Sound Learning 276 93.20 17.63 102.13 14.98 0.51
Listening Vocabulary 277 88.13 18.98 101.99 13.81 0.73
Geometric Analogies 276 88.72 17.33 103.04 14.13 0.83

Note. N = 278; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample size for Word Reading Fluency test was too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

232 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.30. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With a Reading Learning Disability (RLD) and Matched Control Group (continued)

Matched control
Individuals with RLD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Index
Dyslexia Risk Index 265 82.16 14.33 102.46 14.35 1.42
Dyslexia Diagnostic Index 263 79.82 16.29 102.75 14.69 1.41
Linguistic Processing Index 273 83.26 17.01 102.23 14.89 1.12
Reading and Spelling Index 265 80.71 12.97 102.84 14.18 1.71

Composite
Sight Word Acquisition composite 260 80.82 15.93 101.66 13.57 1.31
Phonics Knowledge composite 246 84.12 12.42 103.17 14.47 1.53
Basic Reading Skills composite 276 81.20 14.05 102.15 14.10 1.49
Decoding Efficiency composite 245 82.04 13.65 102.78 13.96 1.52
Spelling composite 276 81.95 15.53 102.56 14.71 1.33
Reading Fluency composite 264 86.12 13.77 102.96 14.71 1.22
Reading Comprehension Efficiency composite 252 83.61 15.88 102.95 14.32 1.22
Phonological Awareness composite 276 88.97 20.08 101.80 15.37 0.64
Rapid Automatized Naming composite 274 81.75 17.24 102.42 14.23 1.20
Auditory Working Memory composite 276 91.84 16.05 101.00 14.57 0.57
Orthographic Processing composite 268 84.54 16.75 101.51 14.59 1.01
Vocabulary composite 268 84.75 21.82 102.38 13.56 0.81
Reasoning composite 276 88.81 18.74 103.23 14.21 0.77
Vocabulary and Reasoning 2 composite 268 87.49 19.12 102.46 14.07 0.78
Vocabulary and Reasoning 4 composite 268 86.37 18.52 103.03 14.20 0.90

Note. N = 278; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample size for Word Reading Fluency test was too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

Validity TOD 233

TOD • W-700M wpspublish.com

 Table 5.31. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With a Language Disorder and Matched Control Group

Individuals with a Matched control

language disorder group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Test
Picture Vocabulary 33 89.82 17.23 98.94 13.05 0.53
Letter and Word Choice 33 87.45 16.15 99.58 14.83 0.75
Question Reading Fluency 30 88.33 16.04 98.28 15.12 0.62
Phonological Manipulation 33 86.70 14.20 100.85 13.50 1.00
Irregular Word Spelling 33 84.94 16.20 97.85 13.52 0.80
Rapid Letter Naming 33 90.09 14.78 101.91 14.53 0.80
Pseudoword Reading 33 87.61 13.96 99.27 12.03 0.84
Word Pattern Choice 33 92.45 13.05 96.76 14.77 0.33
Word Memory 33 92.85 13.22 99.70 14.09 0.52
Picture Analogies 33 92.88 14.57 100.85 15.46 0.55
Irregular Word Reading 33 83.24 16.78 98.00 15.06 0.88
Oral Reading Efficiency 33 82.42 14.55 99.48 14.59 1.17
Blending 33 92.36 22.95 100.15 14.91 0.34
Segmenting 33 92.09 23.34 96.73 13.51 0.20
Regular Word Spelling 33 85.18 16.60 96.94 13.37 0.71
Rapid Number and Letter Naming 33 88.91 15.16 100.67 15.78 0.78
Letter Memory 33 94.12 13.55 100.52 12.54 0.47
Rapid Pseudoword Reading 31 87.61 14.11 99.64 13.73 0.85
Rapid Irregular Word Reading 32 87.53 14.34 99.13 12.74 0.81
Symbol to Sound Learning 33 92.39 11.85 101.33 12.56 0.75
Listening Vocabulary 33 87.91 12.57 95.33 14.20 0.59
Geometric Analogies 33 90.18 15.51 100.97 14.20 0.70

Note. N = 33; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes were too small to include for Word Reading Fluency, Silent Reading Efficiency Grades 1–5, and Silent Reading Efficiency Grade 6–Adult

tests; and for Reading Fluency and Reading Comprehension Efficiency composites.
bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

234 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.31. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With a Language Disorder and Matched Control Group (continued)

Individuals with a Matched control

language disorder group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Index
Dyslexia Risk Index 33 86.76 16.68 98.42 16.49 0.70
Dyslexia Diagnostic Index 33 83.48 16.51 98.82 15.24 0.93
Linguistic Processing Index 33 85.52 14.66 99.48 14.74 0.95
Reading and Spelling Index 33 85.03 15.27 98.42 14.87 0.88

Composite
Sight Word Acquisition composite 32 84.41 16.44 99.34 12.93 0.91
Phonics Knowledge composite 31 87.55 13.06 99.36 13.09 0.90
Basic Reading Skills composite 33 84.58 13.79 98.48 13.83 1.01
Decoding Efficiency composite 30 87.23 14.25 99.97 13.20 0.89
Spelling composite 33 84.27 16.85 97.48 13.56 0.78
Phonological Awareness composite 33 88.45 21.21 98.88 14.68 0.49
Rapid Automatized Naming composite 33 88.36 15.73 101.55 15.62 0.84
Auditory Working Memory composite 33 91.85 14.60 99.97 14.75 0.56
Orthographic Processing composite 33 87.24 15.11 97.67 15.45 0.69
Vocabulary composite 33 86.91 16.40 96.70 14.39 0.60
Reasoning composite 33 90.42 15.22 101.12 14.84 0.70
Vocabulary and Reasoning 2 composite 33 89.58 15.92 99.55 15.17 0.63
Vocabulary and Reasoning 4 composite 33 87.55 14.36 98.67 15.23 0.77

Note. N = 33; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes were too small to include for Word Reading Fluency, Silent Reading Efficiency Grades 1–5, and Silent Reading Efficiency Grade 6–Adult

tests; and for Reading Fluency and Reading Comprehension Efficiency composites.
bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

Validity TOD 235

TOD • W-700M wpspublish.com

 Table 5.32. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Attention-Deficit/Hyperactivity Disorder (ADHD) and Matched Control Group

Matched control
Individuals with ADHD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Test
Picture Vocabulary 112 97.53 13.32 103.50 13.93 0.45
Letter and Word Choice 112 92.88 14.32 102.76 14.10 0.69
Question Reading Fluency 112 94.02 14.79 104.94 14.17 0.74
Phonological Manipulation 118 94.35 13.91 105.14 14.32 0.78
Irregular Word Spelling 118 91.06 15.48 103.31 14.13 0.79
Rapid Letter Naming 117 92.47 14.54 103.99 13.67 0.79
Pseudoword Reading 118 93.62 13.75 102.52 13.10 0.65
Word Pattern Choice 118 96.14 13.86 103.52 14.76 0.53
Word Memory 118 95.04 14.40 102.87 14.37 0.54
Picture Analogies 118 99.49 14.63 103.33 14.56 0.26
Irregular Word Reading 118 93.33 14.92 101.31 13.96 0.53
Oral Reading Efficiency 118 92.99 16.25 103.64 15.29 0.66
Blending 118 96.90 11.13 103.34 14.44 0.58
Segmenting 118 100.04 12.31 100.00 15.40 0.00
Regular Word Spelling 118 91.81 13.99 104.03 14.24 0.87
Silent Reading Efficiency Grades 1–5 60 89.20 16.59 104.15 15.86 0.90
Silent Reading Efficiency Grade 6–Adult 58 96.22 15.60 105.25 14.00 0.58
Rapid Number and Letter Naming 118 94.47 15.81 105.70 14.51 0.71
Letter Memory 118 95.19 13.86 103.22 13.03 0.58
Rapid Pseudoword Reading 118 92.97 14.31 103.70 15.26 0.75
Rapid Irregular Word Reading 118 93.45 14.81 103.23 13.68 0.66
Symbol to Sound Learning 118 99.14 15.34 101.75 14.29 0.17
Listening Vocabulary 118 98.91 13.42 102.99 14.63 0.30
Geometric Analogies 118 94.14 16.09 104.67 13.44 0.65

Note. N = 118; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample size for Word Reading Fluency test was too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

236 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.32. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Attention-Deficit/Hyperactivity Disorder (ADHD) and Matched Control Group (continued)

Matched control
Individuals with ADHD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Index
Dyslexia Risk Index 112 92.46 15.20 104.50 14.69 0.79
Dyslexia Diagnostic Index 112 90.04 15.10 105.21 15.01 1.00
Linguistic Processing Index 117 91.62 14.48 105.48 15.06 0.96
Reading and Spelling Index 112 91.12 14.71 104.35 14.63 0.90

Composite
Sight Word Acquisition composite 118 92.84 15.48 102.91 13.60 0.65
Phonics Knowledge composite 118 92.88 14.43 103.62 14.98 0.74
Basic Reading Skills composite 118 92.75 14.03 102.35 13.77 0.68
Decoding Efficiency composite 118 92.69 14.91 104.10 14.61 0.77
Spelling composite 118 91.17 14.88 104.07 14.46 0.87
Reading Fluency composite 112 94.51 14.41 105.64 16.09 0.77
Reading Comprehension Efficiency composite 112 92.74 15.93 105.79 15.20 0.82
Phonological Awareness composite 118 96.10 11.56 103.19 15.41 0.61
Rapid Automatized Naming composite 117 92.95 15.41 105.50 14.43 0.81
Auditory Working Memory composite 118 93.76 15.44 103.38 14.84 0.62
Orthographic Processing composite 112 92.63 14.42 103.94 14.79 0.78
Vocabulary composite 112 98.00 13.27 103.52 13.98 0.42
Reasoning composite 118 96.42 14.87 104.83 14.48 0.57
Vocabulary and Reasoning 2 composite 112 98.05 13.20 103.86 14.31 0.44
Vocabulary and Reasoning 4 composite 112 96.63 13.82 104.67 14.38 0.58

Note. N = 118; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample size for Word Reading Fluency test was too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

Validity TOD 237

TOD • W-700M wpspublish.com

 Table 5.33. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Autism Spectrum Disorder (ASD) and Matched Control Group

Matched control
Individuals with ASD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Test
Picture Vocabulary 46 92.83 20.05 102.82 11.90 0.50
Letter and Word Choice 46 91.93 18.53 99.92 12.53 0.43
Question Reading Fluency 40 90.88 17.46 103.76 11.86 0.74
Phonological Manipulation 49 86.84 17.41 102.00 15.23 0.87
Irregular Word Spelling 49 89.71 15.86 100.49 14.92 0.68
Rapid Letter Naming 49 89.10 18.13 99.45 13.02 0.57
Pseudoword Reading 49 92.37 15.68 99.69 12.72 0.47
Word Pattern Choice 49 91.86 13.76 99.24 14.82 0.54
Word Memory 49 96.43 16.68 101.98 15.51 0.33
Picture Analogies 49 93.31 15.33 102.41 17.17 0.59
Irregular Word Reading 49 91.41 18.88 100.16 13.73 0.46
Oral Reading Efficiency 48 93.31 18.70 99.02 15.05 0.31
Blending 49 90.27 21.76 100.88 17.27 0.49
Segmenting 49 89.00 19.08 100.37 15.44 0.60
Regular Word Spelling 49 88.94 16.79 101.33 15.04 0.74
Silent Reading Efficiency Grades 1–5 28 86.75 18.82 102.36 14.63 0.83
Silent Reading Efficiency Grade 6–Adult 20 88.60 14.99 103.88 15.46 1.02
Rapid Number and Letter Naming 49 87.90 17.50 98.37 14.45 0.60
Letter Memory 49 94.29 18.45 102.41 12.21 0.44
Rapid Pseudoword Reading 46 91.46 17.01 100.88 14.19 0.55
Rapid Irregular Word Reading 47 91.04 17.10 98.94 14.19 0.46
Symbol to Sound Learning 49 92.43 14.58 102.39 13.44 0.68
Listening Vocabulary 49 89.33 16.09 101.80 13.06 0.77
Geometric Analogies 49 93.86 16.92 100.27 14.67 0.38

Note. N = 49; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes for Word Reading Fluency test and for Reading Fluency and Reading Comprehension Efficiency composites were too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

238 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.33. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Autism Spectrum Disorder (ASD) and Matched Control Group (continued)

Matched control
Individuals with ASD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Index
Dyslexia Risk Index 45 89.69 18.06 101.94 12.55 0.68
Dyslexia Diagnostic Index 45 87.62 17.20 101.35 14.18 0.80
Linguistic Processing Index 49 86.41 16.97 100.69 16.07 0.84
Reading and Spelling Index 45 89.93 16.66 101.35 12.92 0.69

Composite
Sight Word Acquisition composite 47 90.43 18.45 99.71 13.98 0.50
Phonics Knowledge composite 46 92.46 16.51 100.45 14.17 0.48
Basic Reading Skills composite 49 91.49 17.57 99.94 13.22 0.48
Decoding Efficiency composite 45 90.64 18.02 100.06 14.83 0.52
Spelling composite 49 88.94 16.90 101.20 15.32 0.73
Phonological Awareness composite 49 86.18 19.44 100.96 16.34 0.76
Rapid Automatized Naming composite 49 87.08 19.05 98.88 13.78 0.62
Auditory Working Memory composite 49 93.73 19.01 102.31 15.43 0.45
Orthographic Processing composite 46 89.50 18.11 99.55 14.46 0.55
Vocabulary composite 46 88.72 20.30 102.63 12.26 0.69
Reasoning composite 49 92.94 17.24 101.67 15.98 0.51
Vocabulary and Reasoning 2 composite 46 91.74 18.01 103.00 13.97 0.63
Vocabulary and Reasoning 4 composite 46 90.09 17.85 102.27 14.25 0.68

Note. N = 49; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes for Word Reading Fluency test and for Reading Fluency and Reading Comprehension Efficiency composites were too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

Validity TOD 239

TOD • W-700M wpspublish.com

 Table 5.34. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Intellectual Disability (ID) or Developmental Delay (DD) and Matched Control Group

Individuals with ID Matched control

or DD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Test
Picture Vocabulary 33 80.58 17.26 97.18 15.33 0.96
Letter and Word Choice 33 75.94 14.87 96.38 14.66 1.37
Question Reading Fluency 32 72.78 14.71 96.76 16.31 1.63
Phonological Manipulation 34 69.65 16.27 93.29 17.74 1.45
Irregular Word Spelling 34 70.79 16.43 93.50 17.51 1.38
Rapid Letter Naming 34 71.85 14.48 93.94 18.72 1.53
Pseudoword Reading 34 73.38 12.54 92.74 15.37 1.54
Word Pattern Choice 34 82.24 11.96 100.41 14.76 1.52
Word Memory 34 80.94 18.40 94.65 14.15 0.75
Picture Analogies 34 80.56 14.12 94.32 15.51 0.97
Irregular Word Reading 34 71.03 18.42 93.79 18.51 1.24
Oral Reading Efficiency 27 72.78 13.80 94.39 13.31 1.57
Blending 34 76.21 22.90 92.03 18.93 0.69
Segmenting 34 75.06 21.44 90.82 13.98 0.74
Regular Word Spelling 34 72.03 15.56 93.00 15.28 1.35
Rapid Number and Letter Naming 34 76.03 13.42 95.41 17.88 1.44
Letter Memory 34 77.85 18.96 97.79 13.16 1.05
Rapid Pseudoword Reading 33 72.12 11.98 94.21 16.61 1.84
Rapid Irregular Word Reading 34 72.65 14.24 96.61 15.21 1.68
Symbol to Sound Learning 34 79.26 15.61 97.56 17.07 1.17
Listening Vocabulary 34 73.50 16.41 89.35 13.56 0.97
Geometric Analogies 34 80.38 12.67 98.15 13.69 1.40

Note. N = 34; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes were too small to include for Word Reading Fluency, Silent Reading Efficiency Grades 1–5, and Silent Reading Efficiency Grade 6–Adult

tests; and for Reading Fluency and Reading Comprehension Efficiency composites.
bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

240 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.34. TOD-C Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Intellectual Disability (ID) or Developmental Delay (DD) and Matched Control Group (continued)

Individuals with ID Matched control

or DD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Index
Dyslexia Risk Index 32 72.31 14.44 96.30 15.76 1.66
Dyslexia Diagnostic Index 32 63.31 16.50 93.12 17.61 1.81
Linguistic Processing Index 34 65.18 15.12 93.12 17.60 1.85
Reading and Spelling Index 32 70.56 13.18 94.09 16.03 1.78

Composite
Sight Word Acquisition composite 34 68.59 18.18 95.55 17.34 1.48
Phonics Knowledge composite 33 73.70 10.34 93.44 16.14 1.91
Basic Reading Skills composite 34 72.15 13.14 92.88 16.54 1.58
Decoding Efficiency composite 33 69.79 12.92 95.42 16.65 1.98
Spelling composite 34 69.59 16.98 93.03 16.37 1.38
Phonological Awareness composite 34 68.50 19.75 89.79 18.67 1.08
Rapid Automatized Naming composite 34 70.82 13.80 93.91 18.46 1.67
Auditory Working Memory composite 34 77.06 18.54 95.24 15.31 0.98
Orthographic Processing composite 33 72.82 15.65 98.00 14.42 1.61
Vocabulary composite 33 72.67 18.62 92.32 13.97 1.06
Reasoning composite 34 78.35 12.96 95.74 13.63 1.34
Vocabulary and Reasoning 2 composite 33 77.67 13.91 94.68 14.64 1.22
Vocabulary and Reasoning 4 composite 33 74.48 13.93 93.15 13.81 1.34

Note. N = 34; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes were too small to include for Word Reading Fluency, Silent Reading Efficiency Grades 1–5, and Silent Reading Efficiency Grade 6–Adult

tests; and for Reading Fluency and Reading Comprehension Efficiency composites.
bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

Validity TOD 241

TOD • W-700M wpspublish.com

 Other Clinical Groups The adult Learning Disability in Reading sample was
composed of only 16 individuals and thus was too
Some individuals in the TOD-C child clinical sample
small for a reliable analysis. That said, this group
had primary diagnoses that were unlikely to show
showed a DRI mean of 90.81 and DDI mean of 93.06,
meaningful differences based on the TOD. One such
which are in the expected direction.
group is individuals with a primary diagnosis of a
speech or articulation disorder (n = 62). The TOD
tests were not designed to be sensitive to speech or TOD-E Clinical Sample
articulation problems; however, because there is a
Reading Learning Disability
relationship between language and speech, small
differences would not be unexpected. Results show The primary clinical group of interest for the TOD-E
that the effect sizes of the differences in test means consisted of 31 individuals diagnosed with dyslexia
between individuals with a speech or articula- or a learning disability in reading. (Note that some
tion disorder and their matched controls ranged individuals from this group had comorbid clinical
from small to medium (0.02 to 0.62), though most diagnoses and thus are represented in more than
were small. The effect sizes of the mean differences one group.) Table 5.35 shows the descriptive statistics
between indexes and composites also ranged from and effect sizes for the comparisons between this
small to medium (0.02 to 0.59), though most were clinical group and their corresponding matched con-
small as well. trol group. The expectation was that the measures of
reading and spelling would show large effect sizes
The other group of individuals unlikely to show
of the differences between group means, while the
differences on the TOD are those diagnosed with
effect sizes for the measures of linguistic processing
emotional/mood disorder, deaf/hard of hearing,
and vocabulary were expected to be lower. Results
visual impairment, or other health/mental health
support this expectation.
impairment. Comparisons of these 53 individuals
with a matched control group revealed almost all The effect sizes of the differences between group
small effect sizes, with test effect sizes ranging from means for the reading and spelling tests are all
0.01 to 0.42 and index and composite effect sizes large, ranging from 0.91 to 1.21, and those of the
ranging from 0.02 to 0.43. linguistic processing, vocabulary, and reasoning
tests were medium to large, ranging from 0.59 to
These results contribute validation evidence by
1.23. The effect sizes for the DRI and EDDI were also
demonstrating smaller effect sizes for clinical groups
both large, 0.83 and 1.20, respectively, reflecting the
whose primary deficits are not related to reading
validity of the risk and diagnostic scores to differen-
and spelling.
tiate between individuals diagnosed with dyslexia
or a learning disability in reading and those who
TOD-C Adult Clinical Sample were not. Effect sizes of the mean differences in the
index and composite scores measuring reading and
The TOD-C adult clinical sample was composed
spelling skills were also large, ranging from 1.03 to
primarily of individuals unlikely to have specific
1.38, as were those of scores measuring other related
challenges on skills measured by the TOD. This
skills, ranging from 0.94 to 1.14.
includes individuals diagnosed with emotional/
mood disorder, deaf/hard of hearing, visual impair- These results provide further validation for the TOD
ment, or other health/mental health impairment. In by illustrating that the TOD-E scores distinguish well
adults, this also includes individuals with ADHD. between individuals who have dyslexia or a learning
Unlike children with the diagnosis, adults typically disability in reading and those who do not.
have had intervention and/or developed compensat-
ing strategies that make their skill levels somewhat
comparable to adults without an ADHD diagnosis.

242 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Table 5.35. TOD-E Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With a Reading Learning Disability (RLD) and Matched Control Group

Matched control
Individuals with RLD group
Effect
Test/Index/Compositea n Mean SD Mean SD sizeb

Test
Picture Vocabulary 30 82.27 22.37 98.90 12.14 0.74
Letter and Word Choice 30 82.90 14.05 95.68 14.87 0.91
Sounds and Pseudowords 31 83.81 15.01 99.90 12.85 1.07
Rhyming 31 86.81 11.99 100.16 12.65 1.11
Early Rapid Number and Letter Naming 31 81.16 13.30 97.58 16.08 1.23
Letter and Sight Word Recognition 30 84.77 13.61 97.32 12.90 0.92
Early Segmenting 31 89.81 16.21 99.32 11.46 0.59
Letter and Sound Knowledge 30 80.30 13.45 96.58 11.67 1.21

Index
Dyslexia Risk Index standard score 29 81.38 16.23 94.81 15.61 0.83
Early Dyslexia Diagnostic Index 27 80.78 13.68 97.23 14.52 1.20
Early Linguistic Processing Index 31 81.68 15.19 99.06 14.06 1.14
Early Reading and Spelling Index 27 80.63 13.82 96.58 14.50 1.15

Composite
Early Sight Word Acquisition composite 29 82.28 13.93 96.65 15.09 1.03
Early Phonics Knowledge composite 30 80.20 13.20 98.35 13.20 1.38
Early Basic Reading Skills composite 29 80.59 11.85 96.77 14.04 1.37
Early Phonological Awareness composite 31 85.32 15.88 100.32 12.76 0.94

Note. N = 31; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aSample sizes for Word Reading Fluency and Question Reading Fluency tests were too small to include.

bEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

Combined Clinical Group disability), there wasn’t an expectation of a difference

between tests, indexes, or composites of reading and
Due to the smaller number of individuals with a
spelling compared with those of linguistic processing
clinical diagnosis in the TOD-E sample, all clinical
or vocabulary. However, effect sizes of the differences
diagnoses of interest were collapsed into a single
between the means of the clinical group compared
group. This group of 80 individuals had diagnoses
with the matched control were medium to large
of developmental delay, intellectual disability,
(mostly large), ranging from 0.46 to 1.08.
language disorder, autism spectrum disorder, and
attention-deficit/hyperactivity disorder. Table 5.36 This indicates that the TOD-E test, index, and
shows the descriptive statistics and effect sizes for the composite scores distinguish meaningfully between
comparisons between this clinical group and their typically developing individuals and individuals
corresponding matched control group. Because indi- with one or more of the following diagnoses: devel-
viduals with these diagnoses are likely to have skill opmental delay, intellectual disability, language
difficulties across all areas measured by the TOD disorder, autism spectrum disorder, and attention-
(with the possible exception of those with a language deficit/hyperactivity disorder.

Validity TOD 243

TOD • W-700M wpspublish.com

 Other Clinical Groups composite effect sizes ranged from 0.00 to 0.44.
The TOD tests were not designed to be sensitive to
Twenty-nine individuals in the TOD-E clinical
speech or articulation problems; however, because
sample had a primary diagnosis of a speech or
there is a relationship between language and speech,
articulation disorder that was unlikely to show
particularly at younger ages, these differences are
meaningful differences based on the TOD. Compar-
not unexpected. This absence of large effect sizes in
ing their mean differences with a matched control
a group not expected to differ meaningfully on the
group yielded small to medium effect sizes. Test
skills measured by the TOD-E tests contributes vali-
effect sizes ranged from 0.01 to 0.39, and index and
dation evidence for the TOD.

Table 5.36. TOD-E Child Standard Scores: Descriptive Statistics and Effect Sizes for Individuals
With Developmental Delay, Intellectual Disability, Language Disorder, Autism Spectrum Disorder,
Attention-Deficit/Hyperactivity Disorder, and Matched Control Group

Matched control
Clinical group group
Effect
Test/Index/Composite n Mean SD Mean SD sizea

Test
Picture Vocabulary 80 83.68 17.93 97.73 16.87 0.78
Letter and Word Choice 80 88.50 15.37 98.13 15.10 0.63
Word Reading Fluency 23 85.83 18.53 104.75 19.42 1.02
Question Reading Fluency 55 92.04 15.62 99.16 14.93 0.46
Sounds and Pseudowords 80 87.91 15.57 101.10 14.15 0.85
Rhyming 80 86.66 11.66 99.30 14.67 1.08
Early Rapid Number and Letter Naming 80 87.65 15.20 97.40 16.59 0.64
Letter and Sight Word Recognition 80 89.28 15.30 100.13 13.71 0.71
Early Segmenting 78 89.03 12.87 96.54 15.34 0.58
Letter and Sound Knowledge 80 86.64 15.45 97.90 13.65 0.73

Index
Dyslexia Risk Index standard score 78 88.65 16.00 99.69 16.61 0.69
Early Dyslexia Diagnostic Index 76 85.66 14.21 99.73 15.56 0.99
Early Linguistic Processing Index 78 84.08 13.63 97.41 16.14 0.98
Early Reading and Spelling Index 78 87.28 15.64 100.72 15.47 0.86

Composite
Early Sight Word Acquisition composite 80 88.36 15.28 100.10 15.51 0.77
Early Phonics Knowledge composite 80 85.54 16.99 99.84 15.11 0.84
Early Basic Reading Skills composite 80 86.55 15.97 99.39 15.12 0.80
Early Phonological Awareness composite 78 85.19 12.99 97.78 15.54 0.97

Note. N = 80; some comparisons have smaller ns due to missing scores. Means and SDs are expressed in standard score units (M = 100, SD = 15). All
pairs of means differ significantly, p < .001.
aEffect size (Cohen’s d) = control group mean minus clinic-referred group mean, divided by pooled standard deviation.

244 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Predictive Validity on the same individual and thus are higher than
those reported in the Reliability section. The sample
DDI Predictive Validity consisted of individuals with a diagnosis of reading
The predictive power of the TOD-C Dyslexia disability who had all three Rating Scales (Parent/
Diagnostic Index (DDI) was evaluated using binary Caregiver, Teacher, and Self-Rating) completed and
logistic regression analyses with two groups of 261 a matched sample of individuals from the stan-
students each: 1) students with dyslexia, and 2) a dardization sample (for use in logistic regression
matched control group. Two separate regression analyses, described in the next paragraph). As shown
analyses were conducted. The first used four TOD in Table 5.37, the correlations between the three Rat-
tests that operationalize the Simple View of Reading ing Scales are .77, .77, and .81 for the Self-Rating to
(SVR). SVR posits that Decoding x Linguistic Com- Parent/Caregiver, Self-Rating to Teacher, and Teacher
prehension = Reading Comprehension (Gough & to Parent/Caregiver comparisons, respectively. The
Tunmer, 1986); numerous studies have supported correlations between the Rating Scale and the DRI
the robustness of the SVR in explaining reading and DDI scores were all moderate to large, ranging
comprehension (Hoover & Tunmer, 2021). Irregular from −.64 to −.71 (and significant at p < .001).
Word Reading (11C) and Pseudoword Reading (7C) Logistic regression analyses were conducted to
were operationalized as measures of decoding, and determine the ability of the TOD-C Rating Scales
Picture Vocabulary (1S) and Listening Vocabulary to detect clinically significant weaknesses in skills
(22C) as measures of listening comprehension. Stu- associated with dyslexia by predicting member-
dents with dyslexia were 3.44 times more likely to be ship in the group of individuals diagnosed with
predicted as having dyslexia than students without reading disability or in the matched control group.
dyslexia when the SVR scores alone were included Results indicated that each of the TOD-C Rating
in the model. Students with dyslexia were 9.29 times Scales provides statistically significant improvement
more likely to be predicted as having dyslexia than over chance in detecting reading disability status.
students without dyslexia when TOD-C DDI scores The percentages of correct diagnostic decisions
alone were included in the model (Castleman et al., were 77%, 82%, and 83% for the Parent/Caregiver,
2023). This large increase in predictive power from Teacher, and Self-Rating scales, respectively. Thus,
the SVR to the DDI illustrates the robust ability of the the Rating Scales are credible predictors of students
DDI to accurately predict dyslexia. who have a learning disability in reading, and conse-
quently most likely those who have dyslexia.
TOD Rating Scale Predictive Validity
The TOD Rating Scales were developed to provide TOD-E Rating Scales
another means of gathering information in a com- The TOD-E sample included only Parent/Caregiver
prehensive TOD evaluation, but they can also serve and Teacher Rating Scales. The results in Table 5.38
independently to predict the likelihood of dys- are based on individuals for whom both Rating
lexia. Note that the expectation is that correlations Scales were completed. The intercorrelation coeffi-
between the TOD direct assessment tests and the Rat- cient between the Rating Scales was .75. Correlation
ing Scales will be negative because they are scored in coefficients between the Rating Scales and the TOD-S
opposing directions (e.g., the higher the Rating Scale DRI and TOD-E EDDI were moderate, ranging from
score, the greater the difficulty for the individual −.33 to −.55 (and significant at p < .001). Although
being rated). these correlations are smaller in magnitude than
for the TOD-C, they still demonstrate a relationship
TOD-C Rating Scales between the Rating Scales and the DRI and EDDI.
Because of the small number of students identified
Intercorrelation coefficients between the Rating with a reading disability in the TOD-E sample, no
Scales were presented as evidence of cross-form logistic regression analyses like those conducted for
consistency earlier in this chapter in the Reliability the TOD-C were conducted for the TOD-E.
section (see Tables 5.6 and 5.9). The correlations
reported in Table 5.37 are between ratings completed

Validity TOD 245

TOD • W-700M wpspublish.com

 Table 5.37. Correlations Between TOD Rating Scales and Dyslexia Risk and Diagnostic Index Standard Scores: TOD-C

Parent/Caregiver Teacher Self-Rating

Rating Scale Rating Scale Scale

Parent/Caregiver Rating Scale ­—

Teacher Rating Scale .81 —
Self-Rating Scale .77 .77 —
DRI −.71 −.65 −.70
DDI −.69 −.64 −.65

Note. N = 66. Correlations are based on Rating Scales completed for the same individual. DRI = Dyslexia Risk Index; DDI = Dyslexia Diagnostic Index.

Table 5.38. Correlations Between TOD Rating Scales and Dyslexia Risk and Diagnostic Index Standard Scores: TOD-E

Parent/Caregiver Teacher
Rating Scale Rating Scale

Parent/Caregiver Rating Scale —

Teacher Rating Scale .75 —
DRI −.33 −.40
EDDI −.51 −.55

Note. N = 85. Correlations are based on Rating Scales completed for the same individual. DRI = Dyslexia Risk Index; EDDI = Early Dyslexia Diagnostic Index.

Summary

This chapter described the psychometric research convergent validity. Finally, the TOD Dyslexia
undertaken to support the publication of the TOD. Risk and Diagnostic Indexes distinguish typically
Reliability was examined from several perspectives, developing individuals from those with a reading
and the test, index, and composite scores performed disability. Treatment outcome research is needed to
well based on internal consistency and test–retest expand the range of validity evidence for the TOD.
reliability analyses. The Rating Scales showed good Such research should include studies that assess
internal consistency as well as cross-form consis- individuals with language disorders and other
tency and validity. A confirmatory factor analysis related disabilities, before and after intervention.
showed acceptable fit with the theoretical model These studies will help to validate the TOD as an
upon which the TOD was based. Similarly, the TOD integral component of evidence-based assessment
tests correlate in expected ways with other tests of and intervention planning for individuals with
similar constructs, thereby yielding evidence of dyslexia.

246 TOD Chapter 5 Psychometric Properties

TOD • W-700M wpspublish.com

 Glossary of Terms

alphabetic principle: the basic understanding that spoken language is made up of speech sounds
(phonemes) that can be represented by a letter or letter string (grapheme)
associative memory: recall of the connection between two elements, such as letter names and speech sounds
automaticity: the ability to recognize words quickly
connected text: text that can be read continuously as opposed to word lists
decodable text: reading material that includes words with regular sound–symbol correspondences and that
is used to practice the application of common phonic elements
dyslexia: a neurobiological disorder that causes a marked impairment in the development of basic reading
skills, reading rate, and spelling
fluency: the ability to read a text accurately, quickly, and with appropriate expression
grapheme: the letter or letter combination that represents a single speech sound (e.g., the l in lap, the tch
in catch)
lexical: relating to the words or vocabulary of a language
orthographic mapping: the process of assigning individual speech sounds to the letters that represent those
sounds; this process bonds the spelling, pronunciation, and meaning of a specific word in memory and
explains how children learn to read sight words
orthography: how a language is represented in writing, including the spelling patterns and rules for
punctuation and capitalization
paired-associate learning (PAL): learning and recalling the associations between two stimuli, such as a
symbol and a letter or word
phoneme: an individual speech sound (e.g., cat has three phonemes: /k/ /ă/ /t/)
phoneme–grapheme correspondence: the associations between the speech sounds (phonemes) and the
letters representing those sounds (graphemes)
phonemic awareness: hearing and using individual speech sounds in words; it includes activities such as
combining sounds to read a written word (e.g., putting together the sounds /b/, /ă/, and /g/ to form the word
bag) or pulling apart the sounds to spell a word
phonemic manipulation: tasks that involve altering the order of sounds in a spoken or written word
phonics: an instructional reading method for teaching students the relationships between the individual
speech sounds and the letter or letters that represent these sounds and how to apply these sound–symbol
correspondences to reading and spelling
phonological awareness: the umbrella term that encompasses a broad range of tasks that involve under-
standing and using word parts and speech sounds (e.g., rhyming words, combining the two parts of
compound words, counting the number of syllables within words, counting phonemes)
phonology: the rule system that governs the relationships among the speech sounds of a language
prosody: a component of fluency that includes the patterns of stress and intonation in a language
receptive vocabulary: the words that an individual can understand when spoken or read

TOD 247

TOD • W-700M wpspublish.com

 segment: to break apart compound words, syllables, or phonemes of words
sight word: any word that a reader recognizes instantly without needing to use decoding strategies
sound or phonetic spelling: the words are spelled the way they sound even though the correct letter
combinations may not be used

248 TOD Glossary of Terms

TOD • W-700M wpspublish.com