CS6750 Assignment M5

Anna He
annahe@gatech.edu

Abstract—This series of M assignments seeks to investigate and

redesign an aspect of Strava: an exercise tracking and social me -
dia app. In particular, the investigation will deep dive and dis-
cuss the interface while the user is engaged in a physical activity
i.e., running or cycling. The goal of this project is to use findings
from the investigation to improve the efficiency and ease of ac-
cessing Strava metrics during an activity.

1 QUALITATIVE EVALUATION

Overall, 25 participants responded to my 7-question survey on Peer Survey.

The Survey was designed identically as described in the M4 Assignment, in -
cluding using the original paper interface image from M3. If I could make a
pragmatic change, I would have added more short answer response questions
to better understand why some parts of the design were unclear to some re-
spondents. Instead, all my survey questions were multiple choice. There were
no changes between survey responses or sessions; all administered surveys
were identical.

1.1 Raw Results

An image of the paper design for my lock-screen widget prototype was in-
cluded for question 2 and onwards. A copy of this design can be found in the
Appendix.

Q1: Select your age.

15 of the 25 respondents reported being 18-29 years old. 9 respondents reported
being 30-39 years old. One respondent reported being 40-49 years old.

Q2: How clearly does this design resemble a lock screen? I.e. can you tell
whether a smartphone is unlocked based on this screen?
6 respondents responded “Yes, it is very clear.” 12 respondents responded “Yes,
it is somewhat clear.” 4 respondents responded “Neutral, it is neither clear nor

1
unclear.” 1 respondent responded “No, it is somewhat unclear” and two re -
spondents responded “No, it is very unclear.

Q3: What metrics can you identify based on the design?

18 respondents identified “Mileage/Distance.” 20 respondents identified “Time
Elapsed.” 6 respondents identified “Elevation.” 4 respondents identified “Heart
Rate.” 7 respondents identified “Pace.” 2 respondents identified “Power.”

Q4: How clearly are you able to identify the second screen as a map?
13 respondents said “Very clearly.” 8 respondents said “Somewhat clearly.” 4
respondents said “Neither clearly nor unclearly.” No respondents replied
“Somewhat unclearly” or “Very unclearly.”

Q5: How clearly are you able to identify what button(s) to press to stop the
current Strava activity?
12 respondents said “Very clearly.” 10 respondents said “Somewhat clearly.” 3
respondents said “Neither clearly nor unclearly.” No respondents replied
“Somewhat unclearly” or “Very unclearly.”

Q6: Is it clear how to navigate between the first screen with metrics and the
second screen with the map?
8 respondents responded “Yes, it is very clear.” 10 respondents responded “Yes,
it is somewhat clear.” 6 respondents responded “Neutral, it is neither clear nor
unclear.” 1 respondent responded “No, it is somewhat unclear” and no respon-
dents responded “No, it is very unclear.

Q7: How would you compare this lock screen experience to the existing
Strava experience that requires unlocking your smartphone?
4 respondents said “Much improved.” 13 respondents said “Somewhat im-
proved.” 7 respondents said “Neither clearly nor unclearly.” 1 respondent said
“Somewhat worse.” No respondents said “Much worse.”

1.2 Feedback Analysis

The main takeaways are that most respondents (68%) found this interface to be
an improvement over the existing one (Q7), and could navigate the buttons and
between screens well (Q5, Q6)

Overall, I was quite surprised at the large variety of metrics that were identified
in Q3. Not all of these metrics were necessarily included in the paper prototype,
2
which leads to wondering if this is feedback for which metrics the respondents
wished to see included in future iterations.

I was also surprised at the response to Q2, not foreseeing such difficulty to
identify my paper design as a lock screen. I believe this reflects the low fidelity
of the prototype.

Although expected, I was quite happy that all respondents were able to identify
the second interface as a map in Q4. I was equally happy that no one found it
unclear how to stop the Strava activity in Q5. However, I did expect this result,
since it is directly borrowed from the existing Strava interface within the app.

1.3 Prototype Changes

Given the large variety of metrics in Q3, one change for future prototypes
would be to allow customization of the lock screen metrics. The default appli -
cation would use the most common metrics (Time Elapsed, and Mileage/Dis -
tance) but in the Strava app, there would options to exchange those metrics for
Pace, Elevations or Heart Rate. This would provide more flexibility for the user.

2 PREDICTIVE EVALUATION

This evaluation is for a haptic prototype. In my haptic prototype, the user taps
their phone externally twice to pause their activity. The phone will then vibrate
twice if the activity is being paused, or three times if it is being resumed.

2.1 GOMS Model

The GOMS model for pausing the Strava activity can be found in Figure 1.
Whether an actions is an Operator, method, or selection rules have been identi-
fied before details of the action. There are two main methods to accomplish the
goal: the haptic feedback prototype and the existing Strava app interface.

2.2 Efficiency

Overall, the model for the haptic prototype does show some advantages over
the existing interface. The chief advantage is that it has one fewer operator,
making it more efficient. Moreover, there are fewer selection rules within the
method, making the prototype more accessible to novice users.

3
One way to simplify this model is to make it the action accessible from all faces
of the phone, not only the back of the phone. This would simplify the method
by reducing the number of operators to only two instead of three. However, the
potential trade-off is that the Strava activity could be accidentally paused if the
user is double tapping the front of their screen to turn the screen on. A possible
solution is that double tapping the front of screen to turn the screen on is dis -
abled while a Strava activity is in progress, but this lack of consistency could
easily confuse the user. Alternatively, perhaps double tapping will turn the
screen on and pause the activity. Overall, the method becomes more complex.

Figure 1: GOMS Model for Pausing a Strava Activity

3 EVALUATION SUMMARY

3.1 Additional Needfinding

One important question that requires further needfinding is where and how the
user is carrying their smartphone during a Strava activity. Which users carry
their phone in a pants pocket, in their hand, or in a different bag? Importantly,
how accessible is the phone during the activity? These questions are particu -
larly relevant to the haptic prototype since being able to physically access it is
critical. This question is also important to the lock-screen widget if certain users
keep their phone in a bag and cannot readily access the lock-screen at all. Over-
all, further investigation of how users carry their smartphone is separate from
the prototype itself, but necessary to create better prototypes.

3.2 Design Alternatives

As mentioned earlier, one design alternative is to allow customization within

the lock-screen widget prototype. Within the existing prototype, users could
choose which metric to display on the lock screen. This would require design-
ing a new screen for customization or altering the existing settings screens in
the Strava app.

Another design alternative to explore for the haptic prototype is different

points of contact that enable pausing a Strava activity. In this first prototype,
only the back of the phone can pause the activity. However, as discussed in the
Efficiencies section (2.2), it is worth considering whether to allow tapping the
front of the screen as well. From here, there are further possibilities such as tap -
ping the sides of a smartphone or pressing the volume buttons twice instead.
These are new designs that came to mind during this experience that would be
worth exploring in the next iterations of the design life cycle.

3.3 Prototype Revisions

I would have liked to refine the paper interface a bit more to resemble an actual
phone screen widget. I could have used screenshots of my existing smartphone
lock screen and edited screenshots from Strava, instead of my crudely drawn
designs. A slightly higher fidelity design may have helped the participants.

The evaluations were generally favorable to the prototype, so I also hope to be

able to raise it to the next level of fidelity by creating an actual interactive inter-
5
face. The information and stats would remain dummy values, but the ability to
swipe between two screens and to press the pause button would create some
actual interactivity.

For the haptic prototype, I would like to create a very crude physical object that
can make vibrations. This would provide some physical properties to my inter -
face and allow Wizard-of-Oz prototyping.

3.4 New Evaluations

Assuming an interactive version with dummy values of the lock-screen widget

prototype had been created, it would be ready for a true empirical evaluation.
In this evaluation, I would test the length of time it takes users to retrieve cer -
tain information about Strava. This would help support whether this design is
more efficient or not than the existing interface.

For the haptic feedback, the next evaluation would be qualitative interviews
and focus groups after interacting with a Wizard-of-Oz prototype. It would not
yet be ready for empirical evaluation.

6
4 APPENDIX

Figure 2—Strava lock screen widget as it would

appear on iOS and Android Lock Screens.