Psychology 242; Research in PsychologyPaper questions Your teaching assistant (TA) will be able to answer many questions. Please do not hesitate to e-mail me with questions -- I will post good questions & answers on this site. Several discussion groups are dedicated to the paper: be sure to be prepared for them and ask lots of questions if you have never written a paper like this.
|
Key links: |
Where do we get data for our papers?
To
facilitate your papers we will provide you with data during Week 12. You can also make up your own data, or collect real data. If you have real data available please see me to discuss if it is appropriate.
As you can see from
class, statistics can get complex, so keep your data sets simple. Make up
data for your groups by deciding what the scales are that your variables
are measured by, then write out a set of phony numbers for participants' "scores",
then do your analyses.
For example, if your data are attitude ratings, first decide what your rating
scale is. If you decide on a 5 point rating scale, such as the example given
below, then of course all your data will be numbers between 1 and 5.
Research question: Effect of smoking on life span |
|||
Group
1: |
Group
2: |
||
Participant |
Participant's |
Participant |
Participant's |
Put all your data and the calculations you performed for your statistical analyses in an appendix. This appendix will not count against the page total for the paper. The calculations should be similar to those I am giving you in the class hand-outs, or that the book uses in the statistics section. (The formula for the t-test I used in class and for your homework is more straightforward than the formula in the book).
Can I do a correlation study for my paper, or must it be a t-test?
We want all papers to be based on a "true" experiment, where you have an experimental group and a control group. However, if you have your own data or have a really good idea for a correlation study, you can ask me or your TA for an exception to this rule.
Can I get extra credit by doing more analyses?
Yes! You can do an alternate analysis such as a chi square or a correlation as part of your paper. You could, for example, test to see if you have the same number of men and women in the control v. experimental group as part of your methods section. See your TA or me for advice on this if you want to try it.
Is it necessary to include charts, tables, etc?
Yes, if your data or analyses call for it. A table of results -- such as
mean score for groups -- is always very helpful (and saves time/space in
writing!) in reporting research. Look at the paper in the back of your
book for a good example.
Do I need separate measures & procedure sections, or can I
combine them?
Methods is the larger section that covers the nature of participants, study
procedures, data analyses, etc. If your study is complex each of these sections
should be broken out. If you are doing something relatively simple a single
section called "procedures" could suffice. It is usually better
to err on the side of clarity -- break things into more sections if you think
it will help the reader understand where s/he is at any point in the paper.
Look in your book for examples.
How do I form a control group...what is that?
"I have 2 groups of smokers and I want to test for the influence of
smoking on health. How could I design a control group; what would it be?"
Clearly the control group in that design would have to be a non-smoking group.
The choice of a control group depends upon the research question; what are
you "controlling for"? If you are studying the effects of smoking
on, e.g., cardiovascular health, than the control group must be non-smokers.
If you are examining the effect of, say, a certain type of health promotion
advertising campaign on decisions to quit smoking, than both groups would
be smokers: the experimental group would be smokers who are given the ads,
and the control group would be smokers who were not exposed to the ads. The
research groups are always defined by the research question you are asking.
Is there someone available to help me with statistics?
Raise statistical questions during lecture or in discussion groups if possible; if you have
a question many others probably do as well. Your TAs should have office hours
to discuss your statistics. Look at the handouts available through class
or on the class Web site, and look in the book for examples of studies and
statistical analyses.
Do I need an informed consent in my paper?
What if my experiment involves deception?
"My experiment includes the use of deception, where I use relaxation
with physical therapy patients, but do not tell them about that in advance ....I
believe I do not need an informed consent, but rather I need a debriefing statement.
Is this correct?"
You are part correct -- you need to spend time on your debriefing statement,
to ensure that no one leaves the experiment under duress, to bring them back
to some attitude baseline, explain clearly the need for deception and what
you hypothesis was, etc. You always need a signed
informed consent statement if you are collecting names or any other identifying
information. Even if you do not collect names – for, say, an anonymous
survey -- you must have some form of statement (i.e., on a questionnaire...)
stating that the person can leave any time, or refuse to answer any question.
Students must include both an informed consent and a debriefing statement
in an appendix of their paper. These can be short, but must be complete.
Look in your book for examples.
How many references do we need to the literature? Do we need to
include the articles in what we hand in?
You must have at least two references from regular psychology journals. You
certainly are welcome to include other references from books, text books,
popular magazines or reference works such as Wikipedia, but these do not count as your two references. Include these in your references section, as per the
example in the text. You should turn in the abstracts of the two journal
articles with your paper.
| Paper requirement | Paper structure | go to: research citations |
