19 | User Testing

I conducted two user tests with the paper prototypes so that I could get feedback and recommendations before creating a prototype in Figma. The testers first got some brief information about the project before I presented how the test was going to be conducted. The test consisted of five tasks where different features were involved so that I could understand how the different elements are perceived from others.

  • Is the content understandable and intuitive?
  • How is the interaction of specific elements perceived?
  • When there are multiple ways to solve a task, which one is preferred?

My setup during the user testing

The two people testing the prototype are females in their twenties. Both of them are tech-savvy and have used second-hand apps before.

  1. Register a new user (you can make up the information yourself).
  2. Find sneakers for Thea.
  3. Create an ad for a pair of shoes you want to sell.
  4. Change Martin’s shoe size to size 25.
  5. Change your location.

Task 1: Went quickly, understood how to add a child.

Task 2: Went straight to a specific ad on the home screen, which happened to be a sneakers ad. Didn’t realize you could click on „recommended for Thea“ to see more options.

Task 3: Went well.

Task 4: Did it via the home screen.

Task 5: Profile page -> then settings.

Task 1: Went quickly. Understood ad child. No problems

Task 2: Went to explore, and then sneakers, and saw that it was already set to Thea.

Task 3: Created the ad without problems. Fast

Task 4: Went to profile page and clicked on his shoe size

Task 5: Went to profile page and then settings and clicked on location

  • It is not clear that it is possible to click on the “recommended for Thea” to find a more detailed display. Need to make that section look clickable
  • Changing the shoe size of a child can be done two ways and both paths were used by the test subjects.
  • Most tasks were conducted without any problems.

It was slightly challenging to conduct the tests alone. There is a lot to organize when doing tests with paper prototypes and the organizing can sometimes take the focus away from observing the tests. I also wanted to take notes along the way, but it was too challenging to do both that, observe and organize the paper prototypes. Next time I would want to get help from others to take notes and observe.

🩴 🥿 👠 👡 👢 👞 👟 🥾 🩴 🥿 👠 👡 👢 👞 👟 🥾 🩴 🥿 👠 👡 👢 👞 👟 🥾 🩴 🥿 👠 👡 👢 👞

I will work on the findings, implement some changes and create a prototype in Figma for the big reveal in the last blog post.

Findings from User Tests

I recently conducted user tests on my technical prototype. The prototype does not yet work exactly as I vision my final product, but I find it important to involve users at an early stage. I wanted to test the core idea and get valuable feedback for further development of the concept. 

I chose to conduct the user tests with people I interviewed in my research phase. They are women in the age range 23 to 26 and went to music lessons for 1-3 years as children. They have been playing different instruments such as piano, guitar, drums and the clarinet, but never for a long period of time. To learn more about their musical background, read my blog post from January: Key Findings from Interviews

User test 1

Successful melodies: 4

Observations

  • Wants to test all the buttons before starting the game
  • Tries to sing the melody outload right away after hearing it
  • Points with her fingers on the buttons
  • Uses time to think before trying
  • Improved skills after every try
  • Gets frustrated when failing, but always wants to try again
  • Missing a replay button, to hear the melody again without needing to play it

Playing when looking at the serial monitor (printed notes): much easier, higher level of success. Wants to retry every time she fails. 

General feedback

  • Thought it was very fun to play
  • Felt competitive, did not want to quit
  • Found it very annoying to fail
  • In the beginning, she did not understand that every melody was not necessarily containing all the tones.
  • Low quality speaker makes it hard to separate the tones. Suggests using Max 8 instead of the piezo buzzer. 
  • The “incorrect melody”-sound is similar to some of the melody tones. This could be distracting and confusing. 
  • After successfully playing a melody, the next one plays right away. This was too fast for her to prepare for listening again.
  • If I want to upgrade the product and make it even harder, it could be an idea to also implement different rhythms in the melody. Another option is to make the melodies longer (more than four tones). Nevertheless, she states that it was already complicated enough for her. 

User test 2

Successful melodies: 1

Observations

  • Wants to start right away, before I am finished with the explanation
  • States that she is terrible, does not know anything about music theory
  • Struggles, but learns quickly
  • Notices that she is playing the melody wrong, but struggles to point out what the problem is
  • Decides to give up after numerous attempts on the second melody

Playing when looking at the serial monitor (printed notes): easier, but still needs to think a lot and use multiple attempts to succeed. 

General feedback

  • Said it was very fun, would love to play with it all the time as a child. 
  • Very nice way to train your ear. 
  • Thought it would be easier with a higher quality speaker. 
  • Suggested implementing an orange light for better feedback. It would make it easier to understand the number of wrong notes.  
  • Wanted to try again every time she failed, but it was easy to get stuck on one melody and become annoyed. 
  • Hard to hear what is wrong. 
  • Missed a replay button. 
  • Suggested writing the name of the tones on the buttons. Thinks it would be easier to understand the connections. 
  • Suggested removing the resistors to get brighter LED lights.

Conclusion

In general, it was very helpful to test my concept in such an early stage. Their positive feedback motives me to develop the idea further in the future, and their critique made it clear what changes I should make. I already did some adjustments: 

  • Changed the pitch on some of the tones
  • Increased delay between feedback and new melodies
  • Removed resistors from the breadboard
  • Adjusted the printed messages to the serial monitor