Nt1330 Unit 9 Final Project Part 1

Brian Boisjoli Human-Computer Interaction 12/14/16 Final Project Part 1: For this final project, I have made a few changes to my program to make it more challenging, however, I have added in some functioning to make the program more assistive in learning. There were two large pieces I changed in the program.
The first change I made was adding math problems to the program. There are now three levels to each number, easy (shows the hand gesture as well as the digit), medium (shows only the digit), and hard (gives you a math equation to solve). The equation it gives is fairly simple (equivalent to 2-1 or 4+5). As the program continues the numbers rise from 0 to 9. All numbers in the equations (and their answers) are numbers that the
user has already learned. If they sign the equation wrong they will move down a tier to the difficulty prior to the equation they were currently trying to answer. This allows the user to re-learn the numbers they forgot.
I think implementing these three difficulties is a great way to let the user skip around between numbers forcing them to learn as the continue. The second major addition to the program was a progress bar. This progress bar is displayed in a subplot below both the digit being signed and the warmer or colder image. The progress bar has sections that fill up as the user correctly signs the digit on the screen. If the user incorrectly signs the digit, the progress bar will go down. The bar is split into 5 sections, indicating 20% closer to completion each time it goes up. I think with these two additions, the user will have a much easier time both signing the correct digit, and learning the first 9 digits of the American Sign Language.

Part 2: There were multiple places that I needed to alter my code to be able to implement these changes. The first thing I had to do was delete all my previously termed medium level pictures for the digits and replace them with what was at that time hard level . I then created images for each digit containing a math problem using numbers that had been learned by the user at that point. These images were then placed in the program as the new hard level images for each digit. I now needed to adjust the number-being-signed checker for the math problems because the answer to the problem was not always the current digit the user was learning. This was solved with a few if-statements using the counter to see what stage of the program the user was on. If they encounter an image with a math problem, the number-to-sign will be equivalent to the correct answer. The time given to correctly sign the digit was extended for the hard level because the user needs to think for a moment about the math problem before signing the answer. To add the progress bar I drew all six possible progress levels (0%, 20%, 40%, 60%,
80%, 100%) and took screen shots of each to add to the program. I then created a function for each of these to open them when called. I added if statements to the code in two different spots to either increase the progress bar or decrease based on whether the signed digit was correct or incorrect. To allow this to work, I had to adjust the counter that checks how close the user is to learning the current number.
It now decreases by one each time they sign incorrectly. Previously I had it only increased each time they signed correctly and set to zero each time they signed incorrectly. Now, as you use the program, you can get an indication whether you are correctly signing or incorrectly signing the digit and how close you are to learning it.

Part 3: If I had an additional semester to work on this program, there are many things I would do to increase its usability. With 100 additional testers, I would be able to pinpoint the exact moments of the program that users have trouble with.
This would help greatly because having only tested a few people during the report section of the class; we could have missed many of the flaws that the developer (me) may have been blind to.
To conduct user testing in the future I would write a new use case based on all the adjustments I have made since the last use case written and conduct user testing. I would do this with a portion of the 100 testers and make adjustments, then re-write a use case. These portions would be about 10 users. This way, I could begin to hone in on the exact issues with each version of the program and fix them. I would continue this cycle until the program is looking and working with the user’s preferences. The diversity among the testing group would include both right and left-handed people. It would also be nice to conduct use cases on users who are actually deaf and rely greatly on the pictures. This would give an indication as to whether the images I have used for the program are effective. It would also be nice to have both domain experts, as well as users who have no clue what they are signing up for. This would allow a wide range of results to occur.
Measurements made during user testing would adjust based on the current version of the program. However, a few of the measurements are ones that aren’t
directly involved with the program. For example, I could measure how often the user looks at the moderator or has a confused look. User’s facial expressions are often a big indicator as to their thoughts. Looks of confusion, excitement and other emotions will be noted. I could also measure the users overall attitude towards the program’s

effectiveness. Suggestions from users will also be taken and used in correction. With
100 user tests, I believe this program would improve an incredible amount.