In a classroom experimentation be sensitive with the problem of EXTERNAL and INTERNAL validity. One way to avoid this problem is by using a control group. A control group refers to a group of Ss whose selection and experiences are exactly the same as the experimental group except that they do not receive the experimental treatment (Hatch and Farhady, 1982: 18)
Suppose the research is to investigate the effect of grammar correction on the writing skill in ESL students. INDEPENDENT VARIABLE: the amount of correction and the way correction is given on composition error. DEPENDENT VARIABLE: the degree of grammatical accuracy in Ss’ writing samples.
There are FIVE MAJOR classes of research design.
- One-shot case study: Ss are given some treatment for a period of time and at the end they receive some sort of test. X T
- One-group pretest posttest design. T1 X T2
- Intact group design G1 X T1
True experimental design have three basic characteristics (1) control group, (2) the Ss are randomly selected and assigned to the group, and (3) a pretest is administered to capture the initial differences between the groups.
- Posttest only control group
Initial differences between the groups are controlled for by the random selection and random assignment of the Ss
G1(random) X T1
G2 (random) T1
- Pretest posttest control group design
G1(random) T1 X T2
G2 (random) T1 T2
- Time series design T1 T2 T3 X T4 T5 T6
- Time –sample design T1 X T2 – T3 0 T4 – T5 X T6 – T7 0 T8, etc
- Ex-post facto
- Correlational design, a group of Ss may give the data on two different variables T1 T2
- A criterion group design, two group are compared one measure.
- Factorial design, has more than one independent variable/moderator variable