While it is important to master the use of SPSS software to conduct data analysis, it is equally important to ensure quality in the methods used to collect the data analyzed. Recall the familiar adage, "Garbage in, garbage out," and consider that if data is poorly collected, the analysis of that data will also suffer. Think about how the interrelatedness of the hypothesis, data collection method, and statistical analysis impacts research quality.
Consider the hypothesis you have chosen for your dataset from Week:
How might you design an experiment that will effectively collect data for this chosen hypothesis? How will you minimize threats to validity? Will it be a true experiment or a quasi-experiment? Why or why not?
It is important to note that your assignment is to create an experiment and archival data cannot be used in an experiment.
So note, DO NOT use the existing -archival - data (such as the 1980s/1990s from the military dataset).
Your experiment must have four elements:
1. Random selection of your sample subjects from the population AND random assignment to groups (treatment/control)
2. Pretest measurement on the dependent variable score
3. Some type of intervention
4. Posttest measurement on the dependent variable score
EXAMPLE:
Research question: Do incentives increase recruitment of Army personnel?
Hypothesis: A $10,000 signing incentive will significantly increase the number of recruits.
1. Random selection: 30 recruiting stations will be randomly selected from all recruiting stations in the United States. 1/3 will be selected from the east, 1/3 from the midwest, and 1/3 from the west, including Alaska and Hawaii
2. Pretest: Number of recruits will be measured for January, February, and March of the current year
3. Intervention: Recruits will be offered at $10,000 at the participant recruiting stations during January, February, and March of the upcoming year.
4. Number of recruits from the current year will be compared to the number of recruits of the upcoming year using a paired samples t test.