CPSC 344 - 2016/17 W1 WORKED EXAMPLE – 2016/17 W1 MacLean

Lecture 5a Activity Worksheet

To be completed in groups of 4; 5 if need be.

In this activity, you’ll use pre-generated task examples to compare how two different designs intended to
address the human activities described in the task examples.
• Human activity: Time Tracking, as described in three task examples
• Designs to compare: Two web-based time tracking tools
a) Yast: [Link]
Login: cs344@[Link] / icicsx360
b) GetHarvest: [Link]
Login: cs344@[Link] / icicsx360

Part I – Use three task examples to walkthrough & critique a single design (in pairs)
Divide full group of 4-5 into two subgroups.

1. (Individuals): Read:
• Pg. 4 of handout: Reference list of characteristics of good task examples.
• Pg. 2: The three task examples. Think about how they are different.
Then, allocate ONE of the time tracking sites to each pair. Subgroups, log into your site.

2. Walkthrough: Subgroups: Taking perspective of the person in your task example, “walk
through” one site with each of the three task examples. That is, try to use the site to do the things
you think that person wants to do, keeping in mind that person’s knowledge, motivation,
resources, etc.
AT THE SAME TIME,
Critique: Still in subgroups, on pg. 3, list elements or characteristics of each task example which
MADE IT EASIER OR HARDER to assess the site for the person in the task example.
Remember: you aren’t assessing the interface, you’re assessing the task example!
3. Debrief with full class.

Part II – Discuss site support of each TE (in full group)

1. Return to your full group. Each pair take ~3 minutes to demonstrate the time tracking site you
explored to your group mates.
As you do your demo, focus on showing elements that seem most important to your task example
people.
2. As a full group, go through each task example in turn (3-5m each), and discuss which site best
fits the needs of each one. Add your supporting notes to the TE listing on pg. 3.
Also note when you just couldn’t tell, because there wasn’t enough information in the TE.

1
CPSC 344 - 2016/17 W1 WORKED EXAMPLE – 2016/17 W1 MacLean
TASK EXAMPLES
Note: these are not necessarily GOOD TE’s, in every way. Part of your job is to critique them.

Task Example 1 (Jing): Jing works as an engineer for a consulting firm - partly in the office, and
sometimes at home. She is assigned to client projects by her firm; and might work on one to several
different projects in the course of a week. Furthermore, each project involves a number of different
kinds of jobs - e.g., client meetings, design work, analysis and evaluation, report writing. These job
categories tend to be the same for each client.
For billing and accounting purposes, her firm requires her to track the hours she spends by client, by
type of job for each client, and the date of the hours she spends on each of those jobs.
She is not allowed to bill for the time she spends organizing her time tracking, and generally finds it a
real challenge to remember to record what she’s done when she’s concentrating hard on a job and
juggling the other parts of her busy life. Although she needs the income, she is pretty well paid and she
isn’t always too concerned if she misses a few hours but gets paid for most of them.
On Tuesday (for example), Jing works as follows, with a few breaks: Client 1 -- 3 hours on reporting;
Client 2 -- 1 hour on meeting with client, and Client 3 -- 4 hours on design.
At the end of the month, she submits a spreadsheet with all her hours to her firm’s accountant. The
spreadsheet has to have columns and rows according to a format specified by her firm.

Task Example 2 (Winston): Winston is a TA for CPSC 344. He is paid hourly (and needs every
penny!) He has just one paying client - the university department, represented by the course instructor,
who in turn is required to confirm his bi-weekly hours report.
The instructor and department want him to put his work into several categories - e.g. marking,
supporting lecture or lab, preparation. He therefore lists his hours in a word document on any day he’s
put in time for the course.
Keeping track of the time on each task during the day can be tricky. Sometimes it’s in large blocks, but
sometimes just a few minutes here and there when he can find it. He’s in different places when he does
it, and doesn’t always have his laptop with him, so he often uses the clock on his smartphone then writes
it into a notebook or his word document later on when he gets a chance.

Task Example 2 (Laisha): Laisha is a student with a very active social life. She organizes many things
for groups she’s part of, and spends a lot of time on social media and generally connecting with people.
Unfortunately, her studies seem to need more time than she has to give them lately, and her parents are
going to be pretty unhappy about the marks. Since they pay all the bills, this is definitely not a good
thing and is causing her a lot of stress and in fact she is getting worried about how things are going to go
when she sees them at the next holiday, on a long-planned visit.
She wants to figure out where all that time is going! So she’s decided to do an experiment and track her
time without trying to change anything about her activities - just hit a button whenever she switches
from one kind of activity to another (like studying, lecture, transportation, meals, Facebook breaks, sleep
time).

2
CPSC 344 - 2016/17 W1 WORKED EXAMPLE – 2016/17 W1 MacLean

CRITIQUE: TASK EXAMPLES and FIT OF DESIGNS

1. Jing - Part II: Critique task example. Part III: How each site does/doesn’t support Jing.
II. Critique:
• The TE gives a lot of detail on what Jing wants to do. Her motivations are well etched out and the
TE does a good job of exposing some important requirements (must allow Jing to track hours by
client, type of job and the date for the hours or be able to automatically generate a spreadsheet)
Equally, this means that we can surmise the kind of inputs that we expect Jing to make, and is
also explicitly stated (Eg. Client 1 – 3 hours on reporting). This is achieved by a in-depth look at
the work contexts of Jing which is well-done.
• One of the main issues with this prototype is the lack of a cohesive complete job. There are lots
of specific individual goals and requirements that are exposed throughout, but the interface does
not mention what different steps the designer would need to account for. One good thing is the
fact it manages to mention the last action (generating a spreadsheet) but misses some items
inbetween. The only thing that the interface really alludes to is the need to balance alerts (as
Jing often struggles to remember) with the fact that it shouldn't be too intrusive (other
competing priorities, and she doesn't mind missing a few hours)
• The fundamental problem with this TE however, is the general lack of runnability. There's a lot of
motivation given to Jing's actions and what her inputs should look like, but without actually
talking about her actions. (Ideally in present-tense)

III. YAST: III. GetHarvest:

• This interface intially seems to meet a lot of • This interface does better on a lot of the
Jing's requirements that were exposed. It above criticisms (Especially point 2) In
allows her to track time by different projects particular, a lot of control is given over
and clients. However, when looking closer we replicating certain sub-activities across
realise that the interface has some issues. multiple workshops. This added with the
• Here inside a single larger project, we can ability to easily go back and add time-
record time by different clients. However, periods to the sheets makes this a better
from the TE the requirement is the other way interface for Jing.
round. For every client project (as a single • In particular it can generate spreadsheets
entitity), there are a standard set of activities easily, which fullfills one of the requirements
(that are similar across client projects), that mentioned.
Jing needs to track for. This doesn't seemed to
be ideally facilitated by the interface.
• Equally the interface requires you to switch on
the time (as per the landing page) for every
sub-project you are working on. Likely Jing will
add in the timings after working on a project
(~3:30 on design for Client 1), rather than
remembering to switch on the timer at the
start. Again doesn't really help Jing!
• There are other issues as well, but these are
particularly egregious.

3
CPSC 344 - 2016/17 W1 WORKED EXAMPLE – 2016/17 W1 MacLean

2. Winston - Part II: Critique task example. Part III: How each site does/doesn’t support Winston.
Critique:
• Classic first issue is the interface dependence that is a constant feature of this TE (and of a lot of
bad TE's). This would invalidate and bias a lot of potential information that you can draw from it.
You gravitate towards certain solutions based on the information given about smartphones and
notebooks in this example. In particular, this example read as an existing solution, using existing
interfaces rather than a proper task example meant to help plan out a new design solution.
• The interface does a decent job of exposing some requirements and giving useful user
motivations. We know that compared to the previous example with Jing, Winston cannot afford to
miss recording even a single engagement that he might have. Equally we know that he requires a
single client, however with several categories underneath.
• In particular, one of saving graces of the TE is the fact that is does manage to cover a reasonably
complete job. We can surmise that since there are much shorter blocks of work than Jing's
example (...a few minutes here and there), a workflow where Winston switches on a timer when
starting off and switches off later is probably well-suited to this example. Again, this information
should ideally be provided without suggesting an interface specific solution or explaining an
existing one (like this TE)

YAST: GetHarvest:

[Shorter for TE2 and TE3 as a lot of points from [See above.]
TE1 apply here]
• Winston could create multiple 'projects'
(where each stand for a specific activity in
UBC as a TA), and start the timer when he
works on them. This is simple and can be
done even without a laptop (based on having
a smartphone as per the TE). This interface
seems to suite Winston in comparison to
GetHarvest, which also manages to do a lot
of these items but are probably more work
than Winston needs to ideally put in.

3. Laisha - Part II: Critique task example. Part III: How each site does/doesn’t support Laisha.
Critique:

• This TE is confusing as it spends a lot of time motivating the details behind Laisha actions, without
giving enough meat about the actual design. This is a classic TE problem, where extraneous
information becomes the norm, with very few requirements being exposed. There is some
confusion at to what exactly is this interface for: all that is seems like is a simple time-tracker at
the end of the TE.
• In particular, there is some critical information that is provided by this TE, in the final paragraph.
However, this is spoilt by the interface specificity (instead of saying hit a button, you can mention
that she notes the starting and ending time instead. While this seems like a very inconsequential
difference, you have otherwise basically biased your solution already. What if you found out that a

4
CPSC 344 - 2016/17 W1 WORKED EXAMPLE – 2016/17 W1 MacLean
voice-based activation interface was better later for example?
• The TE does a poor job of describing an entire job. It basically just informs us about one potential
input by the user and nothing else (click a button when you start) What about switching off when
you stop? Or is that assumed that since it's another activity, that is done automatically by the
interface? These are the questions that are required to be asked.

YAST: GetHarvest:
[Similar to Winston in terms of Critique. Yast is [Similar to Winston in terms of Critique.]
probably simpler overall for Laisha's
requirements]

5
CPSC 344 - 2016/17 W1 WORKED EXAMPLE – 2016/17 W1 MacLean

REFERENCE: Good task examples . . .

1. Say what the user wants to do but does not say how they would do it
• no specific assumptions made about the interface
• can be used to compare different design alternatives in a fair way
2. Are very specific
• says exactly what the user wants to do
• specifies actual items the user would eventually want to input (somehow)
• tasks can be hierarchical

3. Describe a complete job

• not just a list of simple independent goals!
• forces designer to consider how different steps will work together
• contrasts how information is carried through the dialog
– where does information come from? where does it go?
4. Say who the users are
• design success strongly influenced by what users know
• name names, if possible
• reflect real interests of real users
• helps find tasks that illustrate functionality in a person’s real work context

5. As a set, identify a broad coverage of users and task types

• the typical ‘expected’ user typical routine tasks
• the occasional but important user infrequent but important tasks
• the unusual user unexpected or odd tasks

6. ARE EVALUATED (final and very important step)

• circulate descriptions to users, and rewrite if needed
à ask users for
omissions corrections clarifications suggestions