Unit - I: Introduction: Various methods to ascertain knowledge, scientific method

and its features; problems in measurement in behavioural sciences; levels of

measurement of psychological variables - nominal, ordinal, interval and ratio
scales; test construction - item analysis, concept and methods of establishing
reliability, validity and norms.
Unit - II: Sampling: Probability and non-probability; various methods of sampling
- simple random, stratified, systematic, cluster and multistage sampling;
sampling and non-sampling errors and methods of minimizing these errors.

Here are detailed notes on methods to ascertain knowledge, the scientific method, and its
features, based on standard principles:

1. Methods to Ascertain Knowledge

Knowledge can be acquired through various methods, each with its own advantages and
limitations:

A. Intuition

 Definition: Knowledge gained through a sudden insight or gut feeling without

reasoning.
 Advantages:
o Quick and spontaneous.
o Useful in creative processes.
 Limitations:
o Lacks systematic approach.
o May not be reliable or consistent.

B. Authority

 Definition: Accepting knowledge based on the credibility of a source (e.g., experts,

books).
 Advantages:
o Saves time by relying on established facts.
o Trustworthy when sources are reliable.
 Limitations:
o Susceptible to bias or misinformation.
o Not suitable for novel or evolving knowledge.

C. Rationalism

 Definition: Using logic and reasoning to derive conclusions.

 Advantages:
o Systematic and structured.
o Does not rely on empirical evidence initially.
 Limitations:
o Relies heavily on premises being true.
 o May not always align with empirical findings.

D. Empiricism

 Definition: Gaining knowledge through observation and sensory experience.

 Advantages:
o Based on tangible evidence.
o Promotes experimental verification.
 Limitations:
o Limited by the observer's accuracy and environment.
o Cannot explain abstract concepts without interpretation.

E. Scientific Method

 Definition: A systematic approach combining observation, experimentation, and

reasoning.
 Advantages:
o Combines empirical and rational approaches.
o Reliable, consistent, and adaptable.
 Limitations:
o Time-consuming.
o May require specialized tools or techniques.

2. Scientific Method

The scientific method is a systematic, logical, and empirical approach used to acquire
knowledge and test hypotheses.

Steps in the Scientific Method:

1. Observation:
o Identifying and defining a problem.
o Gathering initial information.
2. Formulating Hypotheses:
o Developing testable and measurable predictions.
3. Experimentation:
o Designing experiments to collect data.
o Controlling variables to ensure accuracy.
4. Analysis:
o Interpreting data using statistical tools.
o Identifying patterns or relationships.
5. Conclusion:
o Confirming or rejecting hypotheses.
o Providing explanations for observed phenomena.
6. Replication:
o Repeating experiments to verify results.
o Ensuring reliability across different settings.
3. Features of the Scientific Method

A. Systematic and Organized:

 Follows a structured sequence of steps to ensure accuracy and clarity.

B. Empirical:

 Relies on observable and measurable evidence.

C. Objective:

 Avoids personal biases or subjective interpretations.

D. Testable:

 Hypotheses and theories can be tested and falsified through experimentation.

E. Reliable and Repeatable:

 Results are consistent when the process is repeated under similar conditions.

F. Self-Correcting:

 Open to modifications based on new evidence or discoveries.

G. Universal Applicability:

 Can be applied across various fields and disciplines.

H. Logical and Rational:

 Uses reasoning to interpret results and draw conclusions.

Summary Table:

Method Definition Example

Intuition Gut feeling, sudden insight Choosing a solution instinctively.
Authority Based on credible sources Referring to textbooks for theories.
Rationalism Logic and reasoning Deductive reasoning to prove a theorem.
Empiricism Observation and experience Conducting surveys to gather data.
Scientific Method Systematic investigation Testing a hypothesis through experiments.

Chat gpt.
Here are concise notes on the problems in measurement in behavioral sciences:

1. Conceptual Problems

 Ambiguity of Constructs: Behavioral concepts such as intelligence, personality, and

motivation are difficult to define precisely, leading to varying interpretations and
measurements.
 Operationalization: Converting abstract constructs into measurable variables can be
challenging and may result in oversimplification or misrepresentation.

2. Reliability Issues

 Inconsistent Measurements: If a measurement tool does not produce consistent

results across different instances or settings, its reliability is compromised.
 Types of Reliability:
o Test-Retest: Consistency of results over time.
o Inter-Rater Reliability: Agreement between different raters or observers.
o Internal Consistency: How consistently the items in a test measure the same
construct.

3. Validity Concerns

 Content Validity: Whether the measurement tool covers the full range of the
construct it aims to measure.
 Construct Validity: Whether the tool truly measures the intended psychological
construct (e.g., measuring depression using the right items).
 Criterion Validity: How well the measurement correlates with a relevant outcome
(e.g., academic performance related to IQ).
 External Validity: The extent to which findings can be generalized to other
populations or settings.

4. Bias and Subjectivity

 Observer Bias: The tendency for researchers or evaluators to unconsciously influence

results based on personal beliefs or expectations.
 Cultural Bias: Tools may not be applicable across diverse cultures or populations,
leading to inaccurate assessments.
 Social Desirability Bias: Participants may provide answers they believe are more
socially acceptable, rather than truthful, affecting the validity of the data.

5. Measurement Tools Limitations

 Poorly Designed Instruments: Inadequate or outdated measurement tools may fail to

capture the complexity of human behavior.
 Standardization Issues: Lack of standardized procedures for administering or
scoring tests can result in inconsistent measurements.
 Saturation Effect: Certain measurement tools may lose their effectiveness after
repeated use or exposure.

6. Sampling and Generalization Problems

  Sample Size: Small sample sizes can lead to unreliable results, failing to represent the
larger population accurately.
 Sampling Bias: If the sample is not representative of the target population, the
findings may not be generalizable.
 Cohort and Temporal Bias: Measurements may not capture differences across
different age groups, cultural backgrounds, or time periods.

7. Ethical Issues

 Informed Consent: Ensuring that participants understand the nature of the study and
the measurement procedures is crucial to avoid ethical dilemmas.
 Privacy Concerns: Safeguarding the confidentiality of participant data to prevent
misuse or harm.

8. Complexity in Multi-Variable Relationships

 Interaction Effects: Behavior is often influenced by a combination of variables,

making it difficult to isolate the effects of one variable.
 Causality vs. Correlation: It can be challenging to determine whether a relationship
between variables is causal or simply correlational.

Chat gpt.

Here’s a detailed explanation of the levels of measurement of psychological variables,

including examples and their unique characteristics:

1. Nominal Scale
 Definition: The nominal scale is the most basic level of measurement. It categorizes
data without any order or magnitude.
 Characteristics:
o Categories are mutually exclusive.

o No inherent order between categories.

o Numerical labels can be used, but they don't indicate quantity or order.

 Examples:
o Gender: Male = 1, Female = 2.

o Types of therapy: Cognitive Behavioral Therapy (CBT), Psychodynamic,

Humanistic.
o Blood group: A, B, AB, O.
  Statistical Analysis: Mode, chi-square test.

2. Ordinal Scale
 Definition: The ordinal scale ranks data in order, but the intervals between ranks are
not equal or specified.
 Characteristics:
o Indicates relative position (greater than, less than).

o Does not specify the magnitude of differences between ranks.

 Examples:
o Severity of symptoms: Mild, Moderate, Severe.

o Ranking in a competition: 1st, 2nd, 3rd.

o Educational qualifications: High school, Undergraduate, Graduate,

Postgraduate.
 Statistical Analysis: Median, rank-order correlation, Mann-Whitney U test.

3. Interval Scale
 Definition: The interval scale provides equal intervals between values but does not
have an absolute zero point.
 Characteristics:
o Equal distances between values.

o Can measure the degree of difference but not the absolute magnitude.

o Zero is arbitrary and not an absolute absence of the variable.

 Examples:
o IQ scores: 100, 110, 120.

o Temperature in Celsius or Fahrenheit: 20°C, 30°C.

o Psychometric test scores.

 Statistical Analysis: Mean, standard deviation, ANOVA, t-tests.

4. Ratio Scale
  Definition: The ratio scale has all the properties of an interval scale, but it also
includes a true zero point, indicating the absence of the variable being measured.
 Characteristics:
o Equal intervals between values.

o Zero indicates a complete absence of the attribute.

o Ratios of values can be meaningfully calculated (e.g., twice as much).

 Examples:
o Reaction time: 0 ms, 50 ms, 100 ms.

o Weight: 0 kg, 10 kg, 20 kg.

o Number of errors made in a test.

 Statistical Analysis: All types of statistical analyses are applicable, including

geometric and harmonic means.

Summary Table

Scale Attributes Examples Statistical Tools

Nominal Categories Gender, therapy type Mode, chi-square

Severity of symptoms,
Ordinal Order Median, rank correlation
rankings

Interval Equal intervals IQ scores, temperature Mean, t-test, ANOVA

True zero, equal All analyses, including

Ratio Reaction time, weight
intervals ratios

Sk mangal notes by chat gpt.

Here’s a detailed explanation of Test Construction and its key components:

1. Item Analysis
Concept:
 Item analysis evaluates individual test items to determine their quality and
contribution to the overall test.
 Helps in identifying items that discriminate well between high and low performers
and those that align with the test's objectives.
Key Metrics:
1. Difficulty Index (P):
o Measures the proportion of examinees who answer an item correctly.

o Range: 0 (very difficult) to 1 (very easy).

o Ideal range: 0.3–0.7.

2. Discrimination Index (D):

o Indicates how well an item differentiates between high and low scorers.

o Range: -1 to +1.

o Items with higher D values are more effective.

3. Distractor Analysis:
o Examines the effectiveness of incorrect options in multiple-choice questions.

o Effective distractors should attract lower performers but not high performers.

2. Methods of Establishing Reliability

Concept:
Reliability refers to the consistency or stability of test scores over time, forms, or raters. A
reliable test produces similar results under consistent conditions.
Methods:
1. Test-Retest Reliability:
o Administer the same test to the same group at two different times.

o Correlate the scores to measure stability over time.

o Suitable for stable traits (e.g., intelligence).

2. Alternate Forms Reliability:

o Administer two equivalent forms of the test to the same group.

o Correlate scores to measure consistency across forms.

o Requires careful construction of parallel forms.

3. Internal Consistency:
o Measures the extent to which items within a test are related.

o Common methods:
  Split-Half Reliability: Divide the test into two halves, correlate
scores, and adjust using Spearman-Brown formula.
 Cronbach's Alpha: Averages all possible split-half reliabilities; widely
used for Likert-scale items.
 Kuder-Richardson Formula (KR-20): Used for dichotomous items
(e.g., true/false).
4. Inter-Rater Reliability:
o Used for subjective tests (e.g., essay scoring).

o Measures agreement between raters using correlation coefficients or

percentage agreement.

3. Methods of Establishing Validity

Concept:
Validity measures whether a test assesses what it claims to measure. It ensures the test's
appropriateness and usefulness.
Types:
1. Content Validity:
o Assesses whether the test covers the entire range of the construct.

o Ensured by expert judgment during test construction.

2. Criterion-Related Validity:
o Correlation between test scores and an external criterion.

o Types:

 Predictive Validity: Test predicts future performance (e.g., SAT for

college success).
 Concurrent Validity: Test correlates with a current criterion (e.g.,
depression test vs. clinical diagnosis).
3. Construct Validity:
o Evaluates whether the test measures the intended theoretical construct.

o Methods:

 Convergent Validity: Test correlates with measures of the same

construct.
 Divergent Validity: Test does not correlate with unrelated constructs.
 4. Face Validity:
o Refers to the appearance of the test being valid to test-takers.

o Not scientifically rigorous but important for acceptance.

4. Establishing Norms
Concept:
Norms provide a reference framework to interpret individual scores relative to a standardized
group.
Steps in Norm Development:
1. Sampling:
o Select a representative sample of the target population.

o Ensure diversity in demographic characteristics (e.g., age, gender, education).

2. Administration:
o Administer the test under standard conditions to the sample group.

3. Statistical Analysis:
o Calculate descriptive statistics (mean, standard deviation).

o Convert raw scores into derived scores (e.g., z-scores, percentiles, T-scores).

4. Establish Norm Tables:

o Create lookup tables for test users to interpret scores.

Summary Table

Component Key Concepts Methods/Tools

Item Difficulty Index, Discrimination

Item statistics
Analysis Index, Distractor Analysis

Test-Retest, Alternate Forms, Internal

Reliability Consistency of test scores
Consistency (Cronbach's Alpha), Inter-Rater

Content, Criterion-Related (Predictive,

Validity Appropriateness of the test
Concurrent), Construct, Face

Norms Reference for score interpretation Sampling, statistical analysis, norm tables
Chat gpt.
UNIT II
Sampling: Probability and non-probability; various methods of sampling - simple
random, stratified, systematic, cluster and multistage sampling; sampling and non-
sampling errors and methods of minimizing these errors.
Here are detailed notes on Sampling in research, covering Probability and Non-Probability
sampling methods along with examples:

Sampling: Definition
Sampling is the process of selecting a subset of individuals or units from a larger population
to represent the entire population in a study. The goal is to make inferences about the
population based on the sample.
Types of Sampling
1. Probability Sampling: In probability sampling, every member of the population has
a known and non-zero chance of being selected. It allows for generalizability and
statistical analysis to estimate the error margin of the sample. There are several
methods within probability sampling:
o Simple Random Sampling (SRS):

 Definition: Every individual in the population has an equal chance of

being selected.
 How it works: Individuals are selected randomly, often using a
random number generator or drawing lots.
 Example: A university wants to survey 100 students about campus
facilities. They randomly select 100 students from a list of 10,000
students.
 Advantages: Easy to implement, unbiased, representative sample.
 Disadvantages: Requires a complete list of the population, impractical
for large populations.
o Stratified Sampling:

 Definition: The population is divided into distinct subgroups (strata)

based on a specific characteristic, and then random sampling is applied
within each subgroup.
 How it works: The strata are usually based on characteristics like age,
gender, income, etc. A proportional or equal number of subjects are
then selected from each stratum.
  Example: A national survey on voting preferences divides the
population into strata based on age groups (18-30, 31-50, 51-70, 71+),
then randomly selects participants from each group.
 Advantages: Ensures representation from all subgroups, improves
precision of estimates.
 Disadvantages: Can be complex to implement, requires knowledge of
population strata.
o Systematic Sampling:

 Definition: A starting point is randomly selected, and then every kth

individual is chosen from a list of the population.
 How it works: The population is arranged in a list, and a sampling
interval (k) is calculated (total population size ÷ desired sample size).
Every kth person is selected after choosing a random starting point.
 Example: A researcher wants to select 100 participants from a
population of 1,000. After randomly choosing a starting point, every
10th person (every 10th name in the list) is selected.
 Advantages: Simple to implement, works well for large populations.
 Disadvantages: If there is a hidden pattern in the population,
systematic sampling might introduce bias.
o Cluster Sampling:

 Definition: The population is divided into clusters (groups), and a

random sample of clusters is selected. Then all or a random sample of
individuals from the selected clusters are surveyed.
 How it works: The population is divided into clusters (geographic
areas, institutions, etc.), and a random selection of clusters is chosen.
Within the chosen clusters, either all members are selected (one-stage)
or random individuals are selected (two-stage).
 Example: A country wants to survey school children. The country is
divided into regions (clusters), and random regions are selected. Then
all schools in those regions are surveyed.
 Advantages: Cost-effective and practical for large populations.
 Disadvantages: Less precise than other methods, as individuals within
clusters may be more similar to each other than to individuals in other
clusters.
o Multistage Sampling:
 Definition: A combination of various sampling methods. It involves
selecting samples in stages, starting with larger units and progressively
narrowing down to smaller units.
 How it works: First, a random sample of larger units (like regions or
districts) is selected. Then, within those units, another random sample
is chosen (like schools within districts), and so on.
 Example: A researcher wants to conduct a national survey on
healthcare. They randomly select a set of cities (first stage), then
randomly select hospitals within those cities (second stage), and finally
select patients from those hospitals (third stage).
 Advantages: Useful for large, geographically dispersed populations,
cost-effective.
 Disadvantages: More complex, can lead to sampling errors at each
stage.
2. Non-Probability Sampling: In non-probability sampling, the chances of selecting
individuals from the population are unknown, which means the sample may not be
representative of the population. It is often used when probability sampling is not
feasible. Common methods include:
o Convenience Sampling:

 Definition: Individuals are selected based on ease of access or

availability.
  How it works: Researchers choose participants who are easiest to
reach or who are most available at the time.
 Example: A researcher surveys people at a shopping mall because they
are easily accessible.
 Advantages: Quick, inexpensive, easy to conduct.
 Disadvantages: Potential bias, not representative of the population.
o Judgmental (Purposive) Sampling:

 Definition: The researcher selects individuals based on their judgment

of who will provide the best insights or information.
 How it works: Participants are chosen because they are thought to
have specific knowledge or experience relevant to the study.
 Example: A researcher studying expert opinions on climate change
may only select environmental scientists to interview.
 Advantages: Useful when specific expertise is required.
 Disadvantages: Potential bias, subjective selection.
o Snowball Sampling:

 Definition: Existing participants recruit future participants from their

social networks, creating a “snowball” effect.
 How it works: Initial participants are selected, and they refer others
who meet the study criteria.
 Example: A researcher studying drug users might start with a few
known individuals and have them refer others in their network.
 Advantages: Useful for hard-to-reach populations.
 Disadvantages: Bias, as the sample may not represent the broader
population.
o Quota Sampling:

 Definition: Researchers select participants non-randomly based on

certain characteristics (age, gender, etc.), ensuring specific subgroups
are included.
 How it works: Participants are selected to meet quotas for specific
characteristics. Once the quota is met for a group, no more individuals
from that group are selected.
  Example: A researcher selects a set number of participants from
various age groups (e.g., 50 people aged 18-25, 50 people aged 26-35)
to match the distribution of the population.
 Advantages: Ensures representation of certain groups.
 Disadvantages: May introduce bias, as individuals are not randomly
selected.

Summary of Sampling Methods

Sampling
Method Advantages Disadvantages
Type

Simple Requires complete list of

Probability Equal chance for all, unbiased
Random population

Ensures representation of Complex, requires knowledge

Stratified
subgroups of strata

Simple, good for large Risk of bias if there’s a pattern

Systematic
populations in the population

Cost-effective, good for large, Less precise, may have

Cluster
dispersed populations similarities within clusters

Suitable for large populations, Complex, potential errors at

Multistage
cost-effective each stage

Non- May not be representative,

Convenience Quick, easy to conduct
Probability biased

Judgmental Useful for expert input Subjective, potential bias

Good for hard-to-reach

Snowball Bias, not representative
populations

Ensures subgroup Potential bias, non-random

Quota
representation selection

These sampling methods help researchers ensure that their samples represent the population
as accurately as possible, although each method has strengths and limitations depending on
the nature of the research and population.
Chat gpt.
Here are detailed notes on Sampling and Non-Sampling Errors and methods of minimizing
these errors:
Sampling and Non-Sampling Errors
In research, errors can arise during the process of collecting data, and these errors can be
classified into two broad categories: Sampling Errors and Non-Sampling Errors.
1. Sampling Errors
Sampling errors occur due to the inherent randomness involved in selecting a sample from
the population. They are the differences between the sample estimate and the true population
parameter. These errors arise because the sample does not perfectly represent the population,
even if the sample is chosen randomly.
Causes of Sampling Errors:
 Sample Size: A small sample size increases the likelihood of sampling error. Larger
samples tend to produce more accurate estimates of the population parameters.
 Sampling Method: The method used to select the sample affects how well the sample
reflects the population. Non-random sampling methods can result in biased samples.
 Sampling Variability: Random variations in which individuals are selected for the
sample can cause differences in estimates between samples from the same population.
Examples of Sampling Errors:
 In a survey about health behaviors, if a researcher only surveys participants from a
gym, the sample might not be representative of the general population's health
behaviors.
 A small sample size can lead to exaggerated or misleading conclusions due to random
fluctuations.
Methods of Minimizing Sampling Errors:
 Increase Sample Size: Larger sample sizes reduce the effect of random sampling
fluctuations and yield more accurate estimates of the population parameters.
 Use Probability Sampling Methods: Employing probability sampling (e.g., simple
random sampling, stratified sampling) ensures that each individual has a known
chance of being selected, leading to a more representative sample.
 Stratified Sampling: This method ensures that distinct subgroups within the
population are adequately represented, reducing the variability that can occur from
random selection alone.
 Use Systematic Sampling: In large populations, systematic sampling can reduce bias
when using a random start, leading to more accurate results than convenience
sampling.
2. Non-Sampling Errors
Non-sampling errors occur due to factors unrelated to the sample selection process. These
errors can occur at any stage of the research process and can affect the accuracy and validity
of the data. They are not due to random variations in sampling but rather issues such as data
collection, data processing, or response bias.
Causes of Non-Sampling Errors:
 Measurement Errors: Errors in the way data is collected. This can occur if the
instrument used for data collection is inaccurate or if questions are misunderstood by
respondents.
o Example: A poorly worded survey question that confuses participants or a
faulty instrument that provides incorrect measurements.
 Non-Response Bias: When individuals selected for the sample do not respond, and
their non-response is not random. This can lead to an unrepresentative sample if the
non-respondents differ in key ways from the respondents.
o Example: A telephone survey on political opinions where people without
landlines (who may have different political views) are excluded.
 Response Bias: Occurs when respondents provide inaccurate or biased responses.
This could happen due to social desirability, where respondents give answers they
think are socially acceptable, or memory recall issues.
o Example: In a survey about substance use, respondents may underreport their
use due to social desirability bias.
 Processing Errors: Errors that occur during data entry, coding, or analysis. These
errors can be caused by human mistakes or technological issues.
o Example: Incorrectly entering data into a spreadsheet or misinterpreting codes
during analysis.
 Coverage Error: Occurs when the sampling frame (the list of individuals from which
the sample is drawn) does not include the entire population of interest. This can
happen if certain groups are excluded from the list.
o Example: A survey about household income that only includes landline phone
numbers, excluding individuals without landlines.
Examples of Non-Sampling Errors:
 If a health survey excludes people living in remote areas, the data may not reflect the
health conditions of those populations, leading to coverage error.
 A self-reported survey on exercise habits may suffer from social desirability bias,
where participants over-report their exercise frequency.
Methods of Minimizing Non-Sampling Errors:
 Improve Survey Design: To minimize measurement errors, questions should be clear,
unbiased, and easy to understand. Pre-testing the survey on a small group before full
deployment can help identify confusing or ambiguous questions.
 Use Valid and Reliable Measurement Tools: Ensure that the instruments (e.g.,
questionnaires, tests) used are validated and reliable. This helps to reduce
measurement errors and ensures consistency in responses.
 Ensure Randomization in Assigning Participants: For experimental studies,
randomization in assigning participants to different groups minimizes bias and
prevents response bias.
 Increase Response Rate: To minimize non-response bias, researchers can use follow-
up reminders, incentives, and multiple modes of data collection (e.g., phone, email,
face-to-face) to encourage participation.
o Example: Offering small incentives for completing a survey can increase
participation and reduce non-response bias.
 Use Multiple Data Collection Methods: Combining self-report methods with
observational or behavioral data can help reduce response bias and improve data
accuracy.
 Data Quality Checks: Implement regular checks and audits during data collection
and analysis to identify and correct processing errors.
o Example: Double-checking entered data for mistakes and ensuring that coding
rules are applied consistently.
 Use Weighting: In cases where certain groups are underrepresented in the sample
(coverage error), statistical techniques like weighting can be used to adjust the results
to account for these imbalances.
o Example: If younger individuals are underrepresented in a survey, their
responses can be given more weight to adjust for this bias.
 Cross-Validation: Cross-validating the data with other datasets or research can help
identify discrepancies and reduce errors.
o Example: Comparing survey results with official government statistics can
help detect systematic biases in the data.
 Training of Survey Administrators: Proper training for individuals who administer
surveys or collect data can help ensure consistency and accuracy in responses and
minimize errors in measurement.

Summary of Sampling and Non-Sampling Errors

Type of Error Cause Example Methods to Minimize

Small sample sizes Increase sample size,

Sampling Random variations in
leading to biased use probability
Errors sample selection
estimates sampling methods

Errors from non-random

Incorrect survey Improve survey design,
Non-Sampling factors (e.g.,
responses, data entry ensure valid and reliable
Errors measurement, response
mistakes instruments
bias)

Misunderstood Pre-test surveys,

Measurement Inaccurate or unclear data
questions, inaccurate improve instrument
Errors collection methods
measuring tools reliability

Participants fail to Not reaching certain Use follow-up

Non-Response
respond, and their non- groups, like non- reminders, offer
Bias
response is biased landline users incentives

Respondents provide Over-reporting of Anonymous surveys,

Response Bias inaccurate answers due to socially acceptable cross-checking
social pressures behaviors responses

Processing Mistakes in data entry or Incorrect coding or data Data validation, regular
Errors analysis entry errors quality checks

The sample frame Broaden the sample

Excluding people from
Coverage Error excludes parts of the frame, use diverse data
rural areas
population collection methods

By understanding and addressing both sampling and non-sampling errors, researchers can
improve the quality and reliability of their data and ensure that their findings are more
accurate and generalizable to the population.
Chat gpt.