Week 1

Week 3

Week 4

Week 5

Week 6

Week 7

Week 8

Week 9

Week 10

Week 2 (May 27 - June 2)

This week Jen and I decided to run a pilot of the survey on a small group of participants, before implementing the survey on large group over several months, in order to work out the kinks of the survey. Most user studies must be approved by a group which oversees the protection of research subjects. At Berkeley, the Committee for the Protection of Human Subjects (CPHS) must approve all research projects which involve human subjects. Our long-term project had been approved by CPHS before my arrival; however, upon proceeding with the pilot, we decided to  inform CPHS of the changes.  Getting permission for the pilot may not have been necessary with CPHS, since technically it's just a smaller version of our already approved Long-Term Survey, but we did it anyway just in case. 

User Study Lesson 2: When dealing with organizations like CPHS, try to think about all the details that would affect your human subjects (e.g. How will data be collected and stored?  How will subjects' identities be protected? What is the payment schedule?) CPHS is concerned with protecting human subjects from disclosure of identity and to make sure they are informed of all terms of the project, including compensation.  (For instance, in our study, as common with many study, we assigned all users with a code number and labeled all data by code number in order to protect their identities).

This week I re-created the necessary CPHS forms, which include a consent form (a contract), a protocol (details of recruitment and data collection), and all related forms that will be used in the survey. CPHS submissions always take much longer than one would think, because survey details must be worked out and related forms must be updated. 

The next step was to select the participants of the pilot. In order to make the data from the pre-selection questionnaire readable, I had to enter the data in a spreadsheet. 

From this data, Jen and I decided to pick a household of 2 persons and a household of 1 person for the trial. We weren't sure what the results of a 2 person household would be in relation to our device, and that's why we picked this type of household for the trial. We also wanted to see what data we would get for a young (30s) single-person household.  Trusting that our actions follow the procedures of the approved Long-Term Survey, we went ahead and contacted our candidates.  We notified the candidates of the possibility of an interview and accompanying activity, but wouldn't proceed with them until we heard from CPHS.

User Study Lesson 3: Set reasonable correspondence deadlines for users. I contacted one potential trial participant and didn't set a clear deadline one when I needed to hear from him regarding his interest. I didn't hear from him for a few days and was in limbo - I didn't know if I should contact someone else or wait. I also wanted to start the study while I was waiting. I should have initially stated that if I don't hear from the user by X date and Y time, I would contact a different candidate. I think a reasonable telephone/e-mail correspondence deadline when dealing with users is within 36 hours to 2 business days.

In light of reading the chapter "Principles and Techniques for Knowing Your Customers" in The Design of Sites: A Pattern Language for the Web (pre-release) by Doug Van Duyne, James Landay, and Jason Hong, I tried to facilitate the process for our users to submit the required data. I had to try to think like the users - think about what they would need to ease and expedite their submission process.  This involved making sure users had all the appropriate materials; they had to have envelopes with the address labels (with our address) and the proper postage (also accounting for the increase in postage by the USPS that would occur in a few weeks).  I also labeled the outgoing envelopes by user code number and week number (so I could easily identify them upon arrival), and pre-stuffed the envelopes with that week's questionnaire.  I had to remember to label the questionnaires with the appropriate user code numbers.  I also thought it would be helpful for the users to have a written schedule so they could keep track of when they mailed their receipts/questionnaires .