Homework 4: Evaluate Designs for Text Entry
Assigned Monday, Feb. 29, 2016, due: Monday, March 14, 2016
Be sure to see the policies for grading, late turn-in, cheating etc, on the homework policy page.
NOTE: TURN IN THIS ASSIGNMENT IN HARDCOPY IN CLASS on Monday, March 14, 2016. Please print out your entire report on paper, staple it together, and bring it to class. (If you have to miss class for any reason, you can turn in your hardcopy in advance in my office, NSH 3517, or if necessary, upload your assignment to Blackboard to demonstrate that you have it done on time.) Be sure your name is on your report!
In this assignment, each student will evaluate three (3) different techniques for text entry using two (2) different people, and write up a report on the results. Note that this assignment is quite similar to Homeworks 1 & 3, but it might take a little longer per participant.
Text Entry Techniques:
Each student will choose three (3) different text entry techniques. We hope that we can get good coverage of all of the kinds of scrolling techniques that are currently in use, with approximately the same number of people doing each one. If you have an "interesting" way of doing text entry, please use that as one of your choices. The assignment of people to techniques will be in this GoogleDoc.
For example, here are some text entry techniques we thought of that you can pick from:
- Typing on a regular keyboard, either on a laptop or a desktop computer.
- Typing on an "ergonomic" keyboard attached to a desktop computer (such as this one).
- Typing on an iPhone or Android on-screen keyboard with one hand with auto-correct on like usual.
- Typing on an iPhone or Android on-screen keyboard with two thumbs with auto-correct on like usual.
- Typing on those phones or tablets with autocorrect turned off. (For example, on the iPhone, you can turn it off from: Settings>General>Keyboard>Auto-correction Toggle switch to Off. For Android: Tap on any text field, such as a search box, e-mail body, or a browser address bar to summon the keyboard, then tap the gear in the bottom left of the keyboard, beside the Sym button, then tap Predictive text to turn it off and on.)
- Using the iPhone or Android on-screen keyboard in landscape mode (instead of portrait mode) -- there might be a difference. Be sure to indicated whether autocorrect was on or off.
- Using Swype on Android phones or iPhone to enter words by swiping across the keyboard
- Using a small physical keyboard that is part of an Android or Blackberry phone (examples: Blackberry classic or Bold, or Droid 3)
- Typing on the on-screen keyboard for a tablet with the tablet sitting on a table or in a holder or a lap (so 2 hands can be used) -- note that all users should use the same pose, whatever you pick.
- Using an attached physical keyboard on a tablet, such as the small keyboards for a Microsoft Surface, or the Logitech keyboard for an iPad (e.g., this one).
- Using Siri or Google voice to dictate the words.
- etc. -- what other text entry techniques can you get access to?
Extra points on this assignment for doing a fourth or more techniques!
Everyone enter the techniques you will do here.
Participants
Every student should test two (2) people:
- Yourself -- you should use the three text entry techniques and measure yourself.
- Someone who is not like you - for example, someone your parent's age, a child, someone with little or no computer skills, etc. It is OK to use the same person you used for "someone not like you" for Homework 1 and/or 3.
Extra points on this assignment for testing a third or more people!
Test Software
We were not able to create special software for this homework, but luckily, there are lots of typing tests on-line. The best one I found is: http://10fastfingers.com/typing-test/english, because it reports both uncorrected errors ("Wrong words") and "Keystrokes" so one can get an estimate of backspaces, and it works on both a desktop and a phone. What I don't like is that you cannot back up to fix previous words, and it uses random words, rather than real phrases. Also, I would have preferred a test that used Wobbrock's analyses, but couldn't find one. Let me know if you find a better typing test than the one above!
Jacob O. Wobbrock and Brad A. Myers. 2006. Analyzing the input stream for character- level errors in unconstrained text entry evaluations. ACM Trans. Comput.-Hum. Interact. 13, 4 (December 2006), pp. 458-489. ACM DL or local pdf. Note: You can skip section 4 -- you do not need to read about the algorithms.
Use the 1 minute test: http://10fastfingers.com/typing-test/english. Each participant should do the test three (3) times, with breaks in between, and you should note if there is any learning effects. Note that you will need to write down the appropriate numbers since the software doesn't save them. You will want to capture for each run the following:
- Words per minute (WPM) - which is "gross WPM", not counting the errors.
- Error keystrokes - which is the red number in parenthesis after Keystrokes - this will approximate the corrected errors
- Wrong Words - which are the uncorrected errors.
You should be able to run the software from any web-enabled device. The software displays the words to type, and then measures how fast and accurately they are entered. Be sure to tell your participants: "Please enter the phrases as quickly and accurately as possible." For your two people, use different orders. (If you do extra techniques or extra people, then still have everyone do all the techniques, with different orders for the different people.) Note that we have designed this homework to take somewhat longer per participant -- about 15 minutes per technique, because we want to measure any learning effects.
Run the software from here: http://10fastfingers.com/typing-test/english
(Please let the Professor know if you find a better test to use!)
Please enter the three times into the class's GoogleDoc form, and also use the results in your own analysis for your report.
Paste your results into this form on GoogleDoc: http://goo.gl/forms/WenPBXj0Zi.
Preference and Demographic Questionnaire
You should create a simple (paper) questionnaire to give each of your subjects (including yourself!) to be filled out at the end after doing all the tasks. It should ask basic demographic information (see below), experience with computers and experience with the devices and text entry techniques you are using. Then, you should have question(s) about their preferences and why. You need to include a blank copy of the questionnaire with your report so we can see what you asked.
Report
The deliverable for this homework is a written report. We expect it to be maybe about 2-3 pages (not including figures or the copy of the questionnaire), single-spaced in TimesRoman 12 point font. Print this report and turn it in before class on the due data. Be sure your name is on your report! The report must cover the following points:S
- A detailed description of your three
text entry techniques. This should be sufficient to allow the reader to reproduce your test exactly. For example, if doing a keyboard on a PC, what is the exact model of computer or keyboard were you using? If a laptop, then what model? If a desktop computer, what kind of keyboard was attached? If doing a Smartphone, what model? It would also be great to include screenshots or photos of your text entry techniques (e.g., a picture of the keyboard).
- A detailed description of your users. (Note: not their names!) Describe all the relevant demographic characteristics, like their experience with computers, their experience with text entry techniques you used, etc.
- A detailed description of the environment in which you did the test. In particular, were the subjects seated at a desk, standing, at a table? If a phone or tablet, was it on the table or in their hands? Did you use the phone in portrait or landscape mode? (You could even test both if you want and see if there is a difference.) What browser did you use to run the test? What test program did you use? Also include the order of the techniques used for each user.
- Then present your results:
- Present the actual measured results for the wpm and uncorrected error rate for each of your text entry techniques for your users and the averages for each technique across the users, etc.
- Discuss the error keystrokes: did people make a lot of errors? Was the error correction method (usually the backspace key) itself hard to enter (did people make mistakes backspacing - note this is not measured by the tool so you will have to observe it or ask the user)?
- Which techniques was the most accurate? Were there particular letters that seemed harder to enter than others? You don't need to report the corrected error rate, but what is your impression of how many errors the participant made and corrected?
- Report on any learning effects. Possibly plot the WPM over the course of the entire study per device. Did people get faster?
- Which did the subjects prefer and why?
- Was there any effect of the order in which people used the techniques?
- Next in your report, have a discussion section that shows that you have command of the following topics from the class syllabus, with respect to the scrolling techniques you studied:
- The design issues regarding the different text entry techniques, especially the trade-offs among them. (e.g. Does a text entry system perform better, but take longer to learn? Is one fast, but at the cost of having many errors? Would one text entry technique benefit from some of the advances in another one? etc.)
- What did your tests show about the performance? Were the tests you used the right ones? What did the tests miss (not measure)?
- Describe the historical progression of text entry techniques and the factors that impacted their evolution and eventual widespread adoption. In particular, compare and contrast your results with the results reported in the Wobbrock and Myers required reading paper.
- Finally, you can optionally have section of future work, for example, if you have any ideas for new studies, or improved ways of doing text entry (for a little extra credit).
- Include a blank copy of your questionnaire (just like it was given to the participants) in an appendix.