Thursday, July 29 Chapter Five Educational Research
1. Systematic sampling: (text pg. 114) every nth member of the population is selected.
In conducting research on parental involvement in education, the researchers used systematic sampling beginning with the 240th person on a random list, followed by 740th, 1240th and so on.
2. Stratified sampling: (text pg. 114) subjects are selected from strata or groups of the population.
The researchers used stratified sampling in order to accurately represent gender in teaching by dividing up teachers into males and females before randomly selecting subjects for their study.
3. Cluster sampling: (text pg. 117) random selection of naturally occurring groups or units and then individuals from the chosen groups for a study.
By choosing subjects from school districts in the state of Washington, the researchers used the method known as cluster sampling.
4. Nonprobability sample: (text pg. 117) the probability of including population elements is known.
In using his Into to Psych class as research subjects, Professor Blindside used a nonprobability sample.
5. Snowball sampling (pg. 121) researcher begins with a few participants and then asks them to nominate or recommend others who are known to have the profile, attributes, or characteristics desired.
Professor Johnny Rockets used snowball sampling in his study on teenagers and their perception of parental involvement by asking his first four subjects to recommend friends with the same point of view.
As we moved into our potential sample for the group projects in Educational Inquiry, the work began to be a little more challenging. Our group worked well together and agreed on our research question and hypothesis. We even discussed our potential sample with ideas in mind. Yet after reading Chapter Five of Educational Research and discussing how to go about determining the sample, things proved to be more challenging.
In order to really see if students in middle school benefit academically from physical fitness, my group felt it was necessary to find students who were not active already. If we simply randomly selected middle school students, we might run into some serious extraneous variables in that many students are involved in organized sports. By eliminating these students from our population, we felt it would be more obvious the impact of a fitness program on academic achievement. In that sense, our study is an intervention. We are intervening in students’ lives to improve their academic achievement through fitness. Yet choosing our sample that would be random enough still remained a difficulty.
After reviewing the options in chapter five, we decided that cluster sampling would be the most effective manner to get the randomness we desire. Since we aren’t interested in income, urban vs. rural, gender or race as a variable, we could really look at schools in each state and choose at random. So that is what we decided to do. We decided upon three schools per state to include in our sample. From there, we will provide a questionnaire to figure out which students do not participate in physical activities and then begin our intervention. In that sense, we are moving beyond the cluster sampling to critical case sampling. These students are