Using statistics in research is only a small part of understanding the whole research story. Whether you are reviewing, using, or creating research studies, understanding key elements of a research study will help you to see how related pieces fit together, and ultimately how statistics can be used to meet research goals. It is important to think through these elements before a study begins to ensure that your results tell you what you really want to know. Consider how the choice of study design could influence other research choices you might make, such as what statistical procedures you need to conduct to understand and interpret your data. Consider also how the choice to measure a variable using a specific scale of measurement can change your results.
For example, consider a study that researches the emotional recovery of individuals after a car accident. Researchers are specifically focusing on age differences in recovery in this study. In this example of a correlational study, the two variables in the study would be age and emotional recovery. If you wanted to look at these variables in another way, you could assign some participants to a therapy group and some to a control group to measure the effect of therapy on emotional recovery. That would be an experimental study. The independent variable would be therapy, and the dependent variable would be emotional recovery. You would not need to consider each participant’s unique mental health status but would need to keep in mind that statistics allow you to consider the group-level differences in data.
This week, by looking at a research scenario, you will be able to identify and differentiate components that make up a study, such as variables, study design, and scales of measurement.
Researchers want to determine if student intelligence is related to self-esteem. They collect a sample of 300 freshmen from a local university’s participant pool and ask each of them to take an IQ test. The results show IQ scores ranging from 95 to 145.
Researchers also ask students to complete a 20-item assessment of their self-esteem. Each item on the survey is scored on a 1 to 5 scale with higher scores meaning higher self-rating of self-esteem. The researchers average the student responses on the 20 items to get a self-esteem composite score.
The researchers conduct statistical analyses to see if it is probable that the sample accurately describes the relationship between the population’s intelligence (as measured with the IQ test) and self-esteem (as measured with the self-assessment questionnaire).
Based on the scenario , submit responses to the following:
- Identify the sample and what population it represents.
- Identify factors to consider when determining what population the sample represents.
- Categorize the variables of self-esteem and IQ based on scales of measurement in the scenario (ordinal, nominal, etc.).
- Identify limitations/factors to consider when evaluating the generalizability of the results.
- State whether this study design is experimental or correlational. Explain your answer.
- Suppose self-esteem was manipulated by telling some students that they were well liked by their classmates, and other students that they were not well liked, before they took an IQ test. If this were the case, identify the independent variable and the dependent variable. Explain how you would know which is which