Psy 359H

Departmental Resources for Psychology Honors Students

Fall 2002

1.  Getting Started:  Selecting an advisor and a research project

      There are some important considerations in choosing an advisor for your research project.  The first is to start the process early.  Faculty members may be busier toward the end of the semester and have less time to help you develop your study.  The second is to try and find a faculty member who has research interests that are a good match with your own.  Consider talking with several faculty members before deciding, to get a better understanding of what the faculty members interests are, and also to clarify and narrow down your own interests.

      There are also several important things to consider when choosing a research idea and designing your study:

a. You may want to keep the project as simple as is reasonable.  There may be many variables that are interesting to you, but simpler studies almost always make more powerful and meaningful contributions because they are much easier to run and interpret.

b. Be cautious of studies that would involve special populations such as the elderly, severely depressed, schizophrenics, or incarcerated persons.  While the research questions that could be addressed using these populations are attractive, you may encounter serious difficulty gaining access to these people. 

c. Longitudinal studies may also present challenges to students who are hoping to graduate.  Keep in mind that it should be possible to complete your study within two semesters.

d. Try to clarify your independent and dependent variables and your hypotheses as soon as you can.  Writing (and rewriting) these early and often will help guide your ideas, design, and interpretation, and thus is beneficial at all stages of the project.

e. Plan your statistical analysis carefully with your advisor.  If you need additional help planning statistics, you may want to seek outside consultation from the sources listed later in this document. 

2.  Library research

      You will probably find yourself involved in library research at many different stages of your project.  In the beginning it is important to read other people’s work because it teaches you about the state of the field and helps you develop ideas for your study.  Once you choose your topic, library research becomes important to help narrow your focus, provide background for your study, and (importantly!) determine that your study has not already been done.  In the later stages of your project research continues to be important—for example, to see how your results fit with those of others, to see what unforeseen implications your work might have, to guide future work that might come out of your study, etc.  Reading articles can also be helpful in deciding out how to write up your own work.

      We at UT are fortunate to have very good library collections and resources.  An obvious place for help accessing all of this material is the reference desk at the libraries themselves.  You can also get help from professors about particular reference resources, important journals, and, of course, previous research papers that relate directly to your topic.  You will almost certainly need to use PsycINFO (http://www.lib.utsystem.edu/ovidweb/ovidweb.cgi) , a web-accessible search engine for finding journal articles and abstracts, and you may also find Medline (a similar resource for more medically/biologically-related work) helpful.  Another useful reference is the Social Science Citation Index, which allows you to identify articles that cite previous articles.  Thus it allows you to search forward: if you have a very good 1992 paper and you want to know what relevant work may have been done since then, the SSCI will tell you.  Some important journals to be aware of (though please note there are many other good ones!) include:  Journal of Abnormal Psychology, Journal of Consulting and Clinical Psychology, Journal of Experimental Psychology: General, Journal of Experimental Psychology: Human Perception and Performance, Journal of Experimental Psychology:  Learning, Memory, and Cognition,  Journal of Personality and Social Psychology, Health Psychology, Psychological Bulletin, Psychological Review, and the Annual Review of Psychology (the last three are sources of review articles.

      Note that the web may be a useful resource at many different stages of your study.  For example, it can be used to “explore” a large topic. This kind of exploration can be useful especially in the early stages when researching the more academic realm seems too intimidating or detail-oriented.  By entering something as general as “trauma and recovery” in a web search engine, wonderful discoveries can be made—you may find sites for relevant societies, journals, books, internet discussion groups, popular magazines, or even personal sites developed to help people locate more information.  Many of these sites provide links to related sites, and so you may get a lot of help in finding the information you need.  For example, there is a site called “Internet Mental Health” (address is http://www.mentalhealth.com/p.html), which has lots of links to sources of information, and even a search engine that allows you to search for journal articles.

      During your research you may come across dissertation citations.  Make every effort to find the published form of this work (e.g., do a PsycINFO search on the author), especially if the dissertation is more than 1-2 years old.  If you cannot find the published version, and if you determine that the dissertation contains important (or the only) background for your topic, you can purchase a copy of the dissertation from a company called UMI (the price is about $30).  Log on to the UMI website for more information (e.g., to search for dissertations, to read abstracts, to order dissertations, etc.).  The address is http://www.umi.com/hp/Products/Dissertations.html.

3.  How do you get your project approved?

      There are three committees that you need to be aware of: the Departmental Review Committee (DRC), the Institutional Review Board (IRB), and the Institutional Animal Care and Use Committee (IACUC).  You will probably need to get approval to conduct your study from either the IRB or the DRC, and if your study involves animals, you will likely need to get approval from the IACUC.

      The Departmental Review Committee (DRC) handles approval for most studies run within the Department of Psychology using human subjects.  Deepa Block in SEA 2.216 (232-4269) has the necessary forms and information for DRC applications.  The application procedure is pretty simple: you will use the forms to submit a proposal of your study, including details about what types of measurements you will take, and you will receive the DRC’s decision shortly thereafter.  Approval of your study MUST be obtained before you start to run subjects. 

      You must seek approval from the Institutional Review Board (IRB) instead of the DRC if your study involves:

      a. Grants for which the student applies.

      b. Any drugs, including over the counter drugs such as aspirin and caffeine.

      c. Children or minors (subjects less than 18 years old). 

      d. Subjecting the participant to more than minimal physical or psychological risk.

Application material for the IRB can also be obtained from Deepa Block in SEA 2.216.  If your project requires IRB approval, you need to consult the departmental IRB representative, Clark Burnham, in SEA 4.312B (phone 475-7129).  It is extremely important that you start the IRB approval process as early as possible.  The IRB typically meets on the fourth Monday of each month, and the deadline for submission of material is 10 days prior to this meeting.  Plan to contact Dr. Burnham well in advance of the deadline, because he may suggest changes to your proposal that should be made before you submit it.  You may not have time if you wait until the last minute.   The good news is that the IRB rarely denies approval flat out.   If there is a problem with your proposal, a “conditional approval” is more likely.  This means that your project will be approved pending certain changes.  When these changes are implemented, you only need to check with Dr. Burnham:  you do not need to submit your application to the IRB again. 

      The third committee is the Institutional Animal Care and Use Committee (IACUC), which reviews and approves projects involving animal research.  Dr. Jerry Fineg (phone 471-7534 or 471-5158) is the director of the Animal Resources Center (ARC) and has final responsibility over project approval.  If you are doing animal research, contact your faculty advisor for the appropriate forms.  Also, see the following website:  http://www.utexas.edu/research/arc/iacuc/iacuc.htm.

4. Where can you get materials for your project?

      After your project has been approved, you can get materials from the business office in SEA 3.214. You can also get scantrons for recording survey data, pencils, multi-media equipment, and various other office-type supplies.

      We have given you a copy code to use for your project (00359); this code will work with the copier in SEA 2.102.  Use of this copier will be closely monitored. Only materials directly related to your research project may be copied.

      You can also make use of the Measurement and Evaluation Center (MEC), located at 2616 Wichita, to score data collected on scantrons and record the data on a floppy disc (your choice of Mac, IBM, or both).  For scanning research forms, the MEC has a $9 setup charge and an additional charge of 4 cents per form.  If scoring is complicated, they are able to write programs to handle the scoring, although you may have to pay an additional fee for this.  You can learn more about the MEC at http://www.utexas.edu/academic/mec/ or call the MEC's scantron scoring office at

471-6508.

5. Where can you get rooms for running subjects?

a. The best location may be space that your advisor has, such as an office or lab, so check with your advisor first.

b.  If you are running groups of subjects and all you need is a regular classroom (e.g. no special equipment like computers or multimedia, etc.), you can sign up for a room in the undergraduate office in the black binder with “rooms” marked on the spine.  This binder is kept on a desk in SEA 2.218.  Before signing up for a room, check with Deepa Block in SEA 2.216 to confirm your experimenter number (this is what you enter in the black binder to reserve a room).  NOTE: rooms are not available at all times, especially at times when a lot of classes are scheduled. 

c.  If you need a room with computers, e.g. for presenting stimuli to subjects in reaction time experiments, several possibilities are SEA 2.114, 2.122, and 2.124.  See Mike Claytor in SEA 2.312H (471-9287) or Bruce Turbeville in SEA 2.312D (475-6835).  Remember, SEA 2.114 and 2.122 is primarily used for PSY 418, so you’ll have to work around that schedule, which can be found on the door to that room or on the Psych Library Server.  Also, SEA 2.124 may be available in the beginning of the semester more than later, because the PSY 458 students may be running more subjects as the semester gets underway.  Remember that the computers in SEA 2.114 and 2.122 use a different password system than the other department computers (except for the instructor’s computer at the front of the room).  So, if you want to use these, see Mike Claytor or Bruce Turbeville to make sure that you and your subjects are able to log on.

d. If you require facilities with multimedia equipment, you can check out a TV/VCR  or tape recorder from the psychology business office, SEA 3.214.  Just sign up on the whiteboard with a dry erase marker.  If you have a very large group, you may be able to get room in the University Teaching Center (UTC) that has built in multimedia equipment in the podium.  For information or scheduling call the scheduling office at 471-6284.

6.  What statistical procedure should you use?

      Some students experience confusion about which statistical procedure to use because it has been a long time since they have studied statistics.  Some good advice might be to review your textbook from PSY 418 or 458 to get a reasonable overview of the various procedures.  It is beyond the scope of this manual to describe statistical procedures in enough depth to be useful; however, the following brief summary of several common statistical procedures may help to get you started. 

      Correlation- A correlation is a relationship between variables, i.e., as one variable changes, another variable changes in a systematic manner. Values of the correlation coefficient (r) range from -1 to +1.  The sign  (+ or -) indicates the direction of the relationship (a negative value means that as one variable increases, the other decreases) and the magnitude  indicates the strength of the relationship (A value of  Zero means that no relationship exists, -1 and +1 indicate that a perfect relationship exists.)

      Regression- Regression uses one variable to predict another, based on the existence of a correlation between the two variables. 

      T-test- This compares the means of two samples to test for a significant difference.  Can be used with independent samples or repeated measures.

      ANOVA (Analysis of Variance) - ANOVAs can be used to test for a significant difference among the means of three or more samples.  ANOVAs include an extensive family of procedures that can be used to test a wide variety of experimental designs for main effects and interaction effects.  If you think you may want to use an ANOVA and aren’t sure which type to use, you should probably read up on ANOVAs and/or consult with someone knowledgeable about statistics for advice.  

      Chi-squared (c2)- This statistic can be used to test for a significant difference between distributions, e.g. between a sample and population, (c2 compares expected and observed frequencies), or between two independent samples.

7.  What statistical programs are available for analyzing data?

      Excel is difficult to use for some procedures, such as ANOVAs, because organizing your data can be complicated.  However, Excel is excellent for computing correlation matrices as long as you do not have any missing data (then it won’t work). If you have missing data, you may want to use SPSS, which will allow you to exclude cases pairwise, permitting you to calculate a correlation matrix in spite of missing data. 

      SPSS is probably the best choice for ANOVAs, t-tests, and chi-squares because organizing the data is straightforward (once you learn how), it’s all menu driven, and you can learn how to organize the data and use the menus by going through the “SPSS tutorial” that is found on most machines in the department.  Reference manuals are in the graduate computer lab on the 2nd floor of the Seay Building.

      LISREL can do model fitting and path analyses.  If you don’t recognize these procedures, you probably don’t need to worry about it.

      SuperAnova can handle almost any imaginable ANOVA design, but it’s probably more powerful and more complicated than you’ll need.                

      Statview is approximately equal in power to SPSS.  You’ll need to find someone to show you how to use it if you don’t know how, or read through one of the manuals in the graduate computer lab on the 2nd floor of the Seay Building.  There is no handy tutorial for Statview.               

      SAS is available in the computer lab on the fifth floor in the Sanchez Building (SZB).  It is especially useful if you need to do confirmatory factor analyses.  These are IBM machines so your Mac discs won’t work here.

8.  Where can you get help with statistics?

      The Student Microcomputing Facility (SMF) offers consulting services for statistics. You can visit their consulting office for a 30 minute appointment, or a consultant can telephone you at a time of your choosing for an appointment by telephone (this will allow you to talk with them while sitting at your computer).   Phone 475-9400 to set up either type of appointment.

      You can also submit questions to the SMF consulting office via email. These usually take one to two working days to resolve. Please be as specific as possible when submitting a question via email. You may email them from this site: http://www.utexas.edu/cc/stat/consult/ehelp-request.html.

      To learn more about statistics help resources, go to this site: http://www.utexas.edu/cc/stat/consult/.