Week 5

1
Week 5

• Everybodys still alive.

2
Housekeeping

• I know it feels odd to do SPSS homework and not
  turn anything in, but thats what were up to.
• Well start turning in syntax files and output
  files after Spring Break.

3
Groves article next week.

• Since most of you skimmed (which is fine) lets
  hold off discussion of Groves until next week,
  when we have time to get through it comfortably.
• I would like you to try to scratch out the design
  of itits more complicated than what we read in
  the intro classmany more conditions.

4
Methods study, due in 3 weeks

• One person so far has checked on a topicId like
  to chat with each of you this week

5
Methodological Study

• Example of a topic
• Youre reading the Petchev piece, and you notice
  the citation of articles about how skip patterns
  affect responses.
• You then remember our textbook with the changes
  in responses based on what the question was next
  to (plagiarism examples).

6

• So you think, Thats interesting stuff. I want
  to read more about question order. And theres
  your study!
• Or, you think, Thats interesting stuff. I
  wonder what other research has been done on the
  techy things you can do in web surveys. And
  theres your study!

7
Journal Evaluations

• Lets talk!

8
Curtin Reading

• Penetration of Caller ID for Southwestern Bell in
  199860
• After 12 calls they contacted the phone company,
  and sometimes kept on callingmakes me think of
  stalker laws. In 1998 they lowered it to 6.
• All of this happening in one month.

9
Lessons from Curtin

• Write down your methodology and put it in a place
  it can be found again and again
• Youll notice they were going back to 1988
  (almost 20 years ago) to examine how they surveyed

10
Lessons from Curtin, II

• Pay attention to the technology available,
  because it might affect your response rateCaller
  ID, for example
• Is there any technology right now that might
  affect web survey response rates?

11
Lessons from Curtin, III

• There is more than one way to calculate a
  response rate
• Some are more conservative than others
• Use the AAPORs rates www.aapor.org
• http//www.aapor.org/rrc.asp

12
Lessons from Curtin, IV

• They found that both noncontacts and refusals
  have increased over the years
• Refusals at about 27
• Noncontacts at about 16

13
Lessons from Petchev

• Their survey included 268 questions. I wonder if
  the effects would be the same on a shorter
  survey.
• Wouldnt it be interesting to conduct a follow-up
  ourselves? Tech is about the same size. Do we do
  any student-wide surveys?

14
Lessons Learned from Petchev

• No sig difference between
• Completion
• Break-off rate and point
• Respondents on ethnicity, gender, year, or credit
  hours

15
Lessons learned from Petchev

• Scrolling took sig longer than paging (but small)
• Sig more people reported using alcohol in the
  scrolling version with links (thinks respondents
  were tempted to skip all the questions)

16

• Scrolling sig higher number of nonsubstantive
  responses (but small)

17

• Sig more students were very likely to complete
  a scrolling version of a survey in a year
• Sig more paging students answered why they
  wouldnt take it with long or time in their
  answers (55 vs. 35)

18
So, the big take-away?

• What does the article tell us for long scrolling
  pages versus short paging ones?

19
Synergy

• Teaching as faculty, most research-intensive
  schools have this load
• 40 teaching
• 40 research
• 20 administration (committees for students and
  the department)
• the tough thing is not letting teaching or admin
  take over everything else

20
When you teach and research the same areas

• We like to say your teaching informs your
  research, and your research informs your
  teaching
• For example, if you are teaching an editing
  course, you can incorporate recent articles you
  found during a lit review for a research project

21
Example from this week

• The articles we read for this week didnt exist
  the last time I taught Quant
• By continually tweaking the class, I found the
  Curtin reading
• Which Ill now cite in my editing survey article
  that had a relatively low response rate

22
Everything builds on each other

• I always imagine it like a tornado that goes up,
  swirling quickly around and around
• Thats when the job is easiest and when its most
  rewarding

23
This week, on the board

• Lets talk about creative ways to increase survey
  response