by Dr. Janet Waters (revised 2017)
An experiment manipulates one or more variables (the independent variable(s), the IV) to determine its effects on another variable(s) (the dependent variable, the DV). You would look to see if there are differences in the DV after the administration of the IV to a group of participants compared to the pre-test DV scores, or between a Control group and an Experimental group. As an example, a researcher hypothesizes that training in interactive imagery production will improve memory for word lists. The imagery training would be the IV, (the variable that is manipulated or varied), and the number of words remembered correctly would be the DV, (the variable that is measured as the effect). In this example, the DV would be measured by simply counting the correct words.
Research Ethics for Experimental Research: As with all research methods, make sure your research proposal has been approved by your instructor or supervisor before conducting your experiment. Your research proposal must include how you plan to gather data on your participants, and a copy of your consent form. Always go over the consent form with your participants before they sign, and ensure their anonymity and confidentiality is protected in your research process (within the legal requirements).
To control for subject & experimenter expectancy effects, you might use a blind or double-blind method in conducting your experiment. For example, you might not want to tell participants the experiment is attempting to increase their memory. Nor would you tell the independent researcher who will code & tally the imagery for the two groups the purpose of the research. The consent form would therefore reflect that you will fully debrief the participants after their participation.
The experiment might be conducted in a lab or in the "field". The DV might be measured by direct observation, or testing of some sort, or even a self-report, as in the imagery example. Try to be as objective and accurate as possible in recording your observations. Use a video or audio recording if necessary and possible, or prepare a checklist of target behaviours ahead of time, & keep accurate records. To control for experimenter bias, you might try to have an independent observer record & code the observations as noted above; ideally, one who doesn't know the purpose of the experiment.
After administering the IV and measuring the effect on the DV, you should have at least 2 sets of numerical (quantitative) data. If you have two different groups of participants, a t-test of the difference between the means (of the DV) can be done. Or an ANOVA (analysis of variance) or f test can be done for 2 or more groups. If you have a single group of participants and are interested in comparing their scores before the IV is administered to their scores after the IV (a pre-test vs post test measure of the same participants), you would use a t-test or ANOVA for "within participants" or repeated measures.
Excel can be used fairly easily to analyze the data. If you have had Psyc 213 (Statistical Methods in Psychology) you should also be able to use R or SPSS to analyze the data.
The standard APA style research report can be used to present the results of your experiment. In the Introduction, briefly review past research & theory in your topic question (e.g. summarize current research on sensory & auditory imagery production). Use APA referencing style to cite your sources. Then in the Method section, present a general description of the groups of participants (number, mean ages, gender, occupation, etc.) in the Participants section, any materials or equipment you may have used in the Materials section, & in the Procedure section, note that your general research strategy was an experiment, & describe your method in enough detail that it could be replicated by someone else.
In the Results section of the report, present your findings. A table and graph should be used to present the means of the 2 or more groups, or of the pre & post test condition. For example,
Memory for Word Lists n = 60
* p < .05
Also note your results in words, & note which statistic you used to analyze the means, & the results you obtained from the statistic. (e.g. "the mean of memory for word lists among the trained imagers was significantly higher than the untrained imagers. (t = ......, p <.05")).
In the Discussion section, relate your results to past or current research & theory you had cited & described in the Introduction. Do note the statistical significance of your findings, & limits to their generalizability. Remember that even if you did not obtain the significant differences you had hoped to, your results are still interesting, & must be explained, with reference to other research & theory.
© Janet Waters, 2017
Capilano University | 2055 Purcell Way, North Vancouver, British Columbia Canada V7J 3H5 Tel: 604.986.1911
Sunshine Coast | 5627 Inlet Avenue, Sechelt, British Columbia Canada V0N 3A0 Tel: 604.885.9310