User Study

User: A[1], a graduate student living in New York. She has no previous experience with any kind of diet tracking applications and currently do not have any plan of losing weight. Her first language is not English.[2]

Procedure: A completed the test remotely, using their own device (iPhone). The observer welcomed them and gave a short introduction of the purpose of the test (“We are doing a usability test on this diet tracking app and in order to improve it, we need you to help us complete 4 tasks”). During the test, the observer walked them through the process and observe their behaviors through laptop camera. After users finishing all 4 tasks, the observer asked if users have any additional comment. Then the observer thanked users for participating the test and concluded the test.

Details of tasks (wording)[3]:

1. Set up your Fooducate account.

2. Find a cookie recipe in the app that will take 30 minutes or less to make?

3. Add either your breakfast, lunch, or dinner today, to the food tracker feature of this app?

4. Find an alternative option to a food item that you scan? It can be any food item with a bar-code.

What they did[4]

Task one: Set up your Fooducate account

The user downloaded the Fooducate application from Apple Store and followed the sign up instruction. The first question was "Are you currently trying to lose weight?". A selected "No". When entering her current weight and height, she was frustrated by the unit. She was only allowed to enter weight and height in lbs and ft. But A is an international student who is more familiar with kg and m. She had to search for a convertor online in order to submit the most precise measurement. When entering gender, A was confused by the wording of this question("Gal/Guy") at first and still able to make the right selection. She said it took her longer to interpret the options since English is not her first language and "Gal/Guy" is not commonly used for registration. The next step is birthdate set up. A reported that the date was automatically set to her actual birthday and she did not know where they collect this information from. Although the birthdate was automatically recorded, A did play around with this function a little bit more. She entered her boyfriend's birthdate on the date selection page and clicked "OK". The date selection page closed and directly jumped to the next question instead of returning to the birthdate page(see below for pictures). She said this step confused her because her birthdate did not appear on the top left corner of the page when she proceed to the next question.

No problem was reported during the rest of registration procedure. After entering required data, an estimated calories intake was calculated. A then linked her account with her email address and created her username. The application also provided some special dietary options (Vegetarian or allergy for example). A do not have any diet restriction so she clicked "done" and was directed to homepage.

Task two: Find a cookie recipe in the app that will take 30 minutes or less to make?

A clicked on the "Recipes" button from the bottom menu bar. Then she clicked the search function and typed "Cookie" in the search bar. The first result is a cookie recipe that takes one minute but this recipe does not have any instruction or pictures. It only contains ingredients. So I instructed A to find another recipe that can really teach her how to make cookies in 30 minutes.

Since the recipe section does not have filters, A had to keep scrolling down and clicked each recipe that claims to be done in 30 minutes. Most of the recipes do not have detailed cooking instruction(oven temperature and cook time). The direction page only has something like "mix it up" or "blah" which is not helpful for non-cooking people just like A. Finally she found a shortbread cookie recipe that takes exactly 30 minutes.

Task three: Add either your breakfast, lunch, or dinner today, to the food tracker feature of this app?

A had one serving of fried chicken and one serving of rice from a Korean restaurant as her lunch.

A first clicked on the "+" button and searched for fried chicken in the database. She first reported that the database only contains packed fried chicken from grocery store. Later she noticed the "FAST FOOD & RESTAURANT" option and scrolled down to look for a data that matches her intake the most. To enter the rice portion, A searched for rice and selected the standard rice data. She entered her intake amount and time, then add them to her tracker.

Task four: Find an alternative option to a food item that you scan? It can be any food item with a bar-code.

A grabbed a bag of Kirkland chocolate almond and scanned it using the food tracker. But the scan function under food tracker did not provider her with alternatives. She was only able to add it into her diet. So A exit food tracker used the scan function on the homepage. This time she was able to look at the nutritious fact of her food and learned about some alternatives.

What was interesting? [5]

• When my users reported that she do not currently have the plan to lose weight, I was surprised that the app did personalized a little bit for her. After the calories calculation page, instead of asking for target weight, target date and calories, A only saw the calories data and was able to set her diet preference. She can always go to her personal information to re-open the weight control function whenever needed.
• When selecting fried chicken from restaurant database, instead of just randomly pick one, my user did worried about not having that specific kind of fried chicken she had. She said "oh there are some restaurant options, but I guess they don't have the one I ordered. It is a Korean place." My interpretation is that she still wanted to make her record as precise as possible even though it is just a test.

What was surprising?[7]

• The user did not struggled about how much food they eat when entering the intake. She just entered everything as one serving although she was not sure how much is one serving and if it is the same with the one serving she had.
• It was surprising that my user had no trouble noticing the "alternative" tag in the food nutritious page. When I test the app, it took me some time to realize existence of menu bar since it is not colored differently from the content.

What they didn't do?[6]

• My users did not complain about the time entering function which is too detailed for me. Compared with other food trackers I used, I feel like the time function is not so helpful and I have to do it for every single item. But in this test, A did the time entry twice without any comment.
• My user did not look for a standard explanation of measurement like "cup" or "serving" although she did not know how much a serving is.
• My user did not enter note for any food item she recorded.
• when completing the second task (cookie recipe), my user did not "naturally" clicked to see detail information(direction) of the recipe she found. I had to check that data and then instructed her to find one with "more information".

Why they did/didn't do things?[8]

• For not complain about the time entering function, first it might because that my user was only adding two items in the test which did not bother her too much. Secondly, it might because that my user do not often have meals on time or she has snacks a lot. So having time entry might work better for her schedule.
• For not looking up for a standard explanation of measurement, the first explanation is the "experiment" setting. I was instructed my users to complete this task so she does not care if the data is precise enough. The second explanation is that she gained the sense of "one serving" from the picture provided. Most data in Fooducate have pictures and users might try to define measurement according to the pictures.
• The experiment setting problem can also explain why A did not use the "add note" function. Also, I personally don't use that function because I only care about nutritious related information about my diet so there is really not much I can "add".
• My user might not noticed the menu bar when looking for recipes so she assumed that what in front of her is everything. At the beginning of task 3, she did not noticed the menu bar for food database at first. Another reason is that the task only requires her to "find a recipe" instead of "a recipe with enough information".

Relation to Readings[9]

• User interview: the article "User Interviews: How, When, and Why to Conduct Them" mentioned some tips about conduction user interviews. I found it to be really helpful when I was trying to ask questions at the end of my usability test. Having a user interview right after the test really helped me to collect information about the internal thoughts of users. I asked users about their definition of "cup" and "serving" to understand my observational data better. This kind of interview has to be conducted right after the usability test, so I was not able to come up with all the questions I want to ask. When writing this report, there are actually more questions I am curious about. But my user might already forgot all the details.
• Contextual Inquiry: "The Role of Observation in User Research" introduced contextual inquiry which can be used as a complement for usability test. I found it could be really helpful when testing applications like Fooducate. As mentioned in previous sections, the "experimental" setting stopped users from acting naturally. So if allowed, conducting a contextual inquiry in addition to usability test can offer more information.

Redesign Speculations[10]

• Allowing more units conversion (meter, kilogram, etc.)
• Making menu bars more noticeable
• Add explanation of measurements (cup, serving)
• Making time entry optional