Basic Between Subjects Designs ANOVA Instructor Marci Smith
Basic Between. Subjects Designs (ANOVA…) Instructor- Marci Smith mercury 1@uab. edu
Between-Subjects Design Experimental design- general structure of the experiment. n Between-subjects design- a design in which different subjects take part in each condition of the experiment. n Within-subjects design- a design in which each subject takes part in more than one condition of the experiment. (Repeated measures) n
Selecting & Recruiting Subjects n The more the sample resembles the whole population, the more likely it is that the behavior of the sample mirrors that of the population.
Practical Limits n n People are often reluctant to participate: ¨ It will take too much time. ¨ It will uncover private information. How to get people to participate: ¨ Make it interesting, nonthreatening, and meaningful. ¨ Emphasize the responsibility of people to aid in research that can help others. ¨ Point out that lots of people do it. ¨ Pay them, if possible.
How many subjects? n n n Too small a sample leads to false results. However, it might not be feasible to run a large number of subjects, due to time or expense. A general rule of thumb is to use at least 15 -20 per group; 30 per group is the most common. If you have 3 groups Medication 1, Medication 2, No Medication ¨ 30 subjects per group minimum 90 subjects total ¨ Probably closer to 100 recruited, to try to offset dropout… ¨
Effect size n Effect size- statistical estimate of the size or magnitude of the treatment effect. ¨ The larger the effect size the fewer subjects needed to detect a treatment effect. ¨ Strong effect size- 10 -20 subjects/group ¨ Moderate effect size- 20 -30 subjects/group ¨ Weak effect size- 30+ subjects/group
One Independent Variable Two Group Design n Two group design- the simplest experimental design, used when only two treatment conditions are needed. ¨ Two independent groups ¨ Two matched groups
Two Independent Groups n Randomly selected subjects are placed in each of two treatment conditions through random assignment. ¨ Random n Selection! Random assignment- every subject has an equal chance of being placed in any of the treatment conditions.
Two Independent Groups Experimental Group- Control Group Design n n Experimental condition- we apply a particular value of our IV to the subjects and measure the DV. Example- Medication group ¨ Experimental group- subjects in an experimental condition. Control condition- used to determine the value of the DV without an experimental manipulation of the IV. Example- No medication group ¨ Control group- subjects in a control condition.
Two Independent Groups Two Experimental Groups n Two experimental groups design- design in which two group of subjects are exposed to different levels of the IV. ¨ Example- 10 mg medication group and 20 mg medication group
Treated Group
Two Independent Groups n When do we use this design? ¨ When there is only one IV ¨ When the hypothesis can be tested with only two treatment conditions.
Two Matched Groups n Two matched groups design- an experimental design with two treatment conditions and with subjects who are matched on a subject variable thought to be highly relevant to the DV. ¨ Common examples- weight, age, IQ, etc…
Two Matched Groups n Precision matching- members of the matched pairs have identical scores/values/etc… ¨ These are usually values that are easily matched. ¨ Example- years of education
Two Matched Groups Range matching- members of a pair fall within a previously specified range of scores. n Choice of range is arbitrary, but the smaller the range the more similar the subjects are. n
Two Matched Groups n Rank-ordered matching- subjects are rank ordered by their scores on the matching variable, subjects with adjacent scores then become a matched pair. ¨ 5. 3, 5. 1, 4. 7, 4. 6, 4. 5, 4. 2, 3. 9
Two Matched Groups n When do we use this design? ¨ You need to eliminate a possible confounding/extraneous. ¨ You need to make it easier to detect the effect of the IV (you have a weak effect). ¨ Example- Decision making study in obese subjects match on education or IQ
Multiple Groups n Multiple groups design- a design with one IV in which there are more than two treatment conditions. ¨ Multiple independent groups design ¨ Matched groups design- possible, but not common because it can be so difficult to match subjects.
Multiple Groups n Multiple independent groups design- the most commonly used multiple groups design in which the subjects are assigned to the different treatment conditions at random.
Multiple Groups n Block randomization- process of randomization that first creates treatment blocks containing one random order of the conditions in the experiment; subjects are then assigned to fill each successive treatment block.
Multiple Groups n Pilot study- a mini-experiment using only a few subjects to pretest selected levels of an independent variable before conducting the actual experiment. ¨ Many grant application require pilot data for new research.
Multiple Groups n When do we use this design? ¨ When you need more than two treatment conditions to test an experimental hypothesis. ¨ When you are trying to detect more complex patterns. n Example- low performance at low & medium levels of IV, but high performance at high levels of IV.
Class Activity
- Slides: 25