Designing Research Exercise



Designing Research Exercise

In both class and laboratory you have been working to understand the interrelationships among research hypotheses, research designs and the validity of the information obtained. So far we have focused on the critical review of research completed by others (i.e., the evaluative side of research design) with a few forays into the challenges of proposing research designs and procedures (i.e., the generative side of research design). This exercise will focus on the latter -- proposing "best" research designs to answer empirical questions.

Study #1 Comparing the benefits of Dyadic Group Therapy for Anxiety and Schizophrenia patients

For the last several years I have been developing and promoting my Dyadic Group Therapy (DGT) for the treatment of persons with social anxiety. The treatment involves twice-weekly therapy sessions for each patient for 16 weeks. There is one therapist and 4 patients in each group. The first session each week is the Dyadic Therapy Session that involves a meeting of just the therapist and an individual patient (the therapist meets each of the four patients in the group at different times). During the 30-minute Dyadic Therapy Session the patient answers a scripted set of questions posed by the therapist (the topics shift each week according to a specific order that is detailed in the DGT Handbook). The second session each week is the Group Therapy Session that involves a meeting of the therapist and all four patients in that group. During the 90 minute Group Therapy Session the patients take turns answering the same questions that they answered in the Dyadic Therapy Session. The idea is that during the Group Therapy Sessions patients will be familiar with the questions and the therapist, and can concentrate on getting used to talking in front of and with other people.

Last month I learned that two of my colleagues (who practice together) were using DGT to treat patients diagnosed with mild-to-moderate schizophrenia. Their idea is that the Dyadic Therapy Session provides "social skills practice" that will then be applied during the Group Therapy Session. They have conducted some research with their patients and claim that the DGT therapy is beneficial.

My questions is whether DGT works better for patients with one of these diagnoses versus the other? Specifically, my hypothesis is that being treated for Anxiety using DGT will produce higher scores on the General Psychiatric Wellness Index (GPWI) than will being treated for schizophrenia using DGT.

Please design an empirical test of my research hypothesis. I have arranged the cooperation of my clinic and that of my colleagues. You have access to our patients. We will deliver the DGT treatment using the standardized procedure to the patients you indicate. We will have the GPWI administered and scored by an independent testing service on the date(s) that you specify.

Consider a few things to get you started…

• Is this a causal or an associative research hypothesis

• What is the IV (proposed causal variable) and DV (proposed outcome variable)

• Can a true experiment be designed and conducted to test this research hypothesis

Propose a research design to answer this question. On the back of this page provide details of participant selection, assignment, treatment provision, controls to provide internal validity, and how the decisions might influence external validity (don't worry about measurement or statistical conclusion validity for this exercise). Details and specifics are the goal -- use of correct terminology and procedural names will be richly rewarded!

Study #2

Below are two causal research hypotheses. Each could be tested using a variety of research designs.

"Squeaky-Clean"

I've just produced a hand cleanser designed to compete with ZEP cleanser that is wildly popular among automotive mechanics. I am certain that Squeaky-Clean can clean that nasty combination of oil, grease and dirt that covers mechanic's hands better than ZEP. I've obtained a spectroanalyzer that you can use to get a good measure of hand "cleanliness", if you want to use it. Mechanics from 25 surrounding garages (about 90 mechanics in all) have been contacted and have agreed to participate in the research.

"Pooch Pals"

For years I've had the best dog training program in the Midwest -- proudly called Doggy Delight. Recently, a new training agency has moved into the state (Pooch Pals). I'd like to be able to advertise that my program works better than theirs, based on an empirical study. Of course, I expect that my program will produce better behaved puppies than theirs. Pooch Pals has agreed to the comparison, following whatever procedures you design. Each of us will provide lists of persons who have inquired about puppy training (about 300) and we will each provide our trainers to provide the actual classes either at our facilities or ones you provide, if you choose to do it that way. If you would like to -- the performance of each trained animal will be assessed by the Nebraska Kennel Association.

For the selected topic, complete the following…

Provide a detailed description of how you would conduct this research as a laboratory experiment. Provide details of participant selection, assignment, treatment provision, controls to provide internal validity, and how the decisions might influence external validity (don't worry about measurement or statistical conclusion validity for this exercise). Details and specifics are the goal -- use of correct terminology and procedural names will be richly rewarded

Provide a detailed description 0of how you would conduct this research as a field experiment. Provide details of participant selection, assignment, treatment provision, controls to provide internal validity, and how the decisions might influence external validity (don't worry about measurement or statistical conclusion validity for this exercise). Details and specifics are the goal -- use of correct terminology and procedural names will be richly rewarded

Provide a detailed description of how you would conduct this research using a nonexperimental design. Provide details of participant selection, assignment, treatment provision, controls to provide internal validity, and how the decisions might influence external validity (don't worry about measurement or statistical conclusion validity for this exercise). Details and specifics are the goal -- use of correct terminology and procedural names will be richly rewarded.

-----------------------

Assignment grade

out of 40 points ___________

Grading

Study #1 ______ 20

Study #2 ______ 20

Total Graded Points ______ - ______ (points lost - why?)

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download