Write your own practice exam question to help yourself (and other students) get more practice. Use one of the datasets weve worked with (like the cal_mega_dataset) to create a practice exam. Heres an R script you can use as a template (Links to an external site.). Post your practice exam to the bCourses discussion if you are feeling groovy about it.
Problem 1 : Describe the Dataset and Research Question (no R needed! Choose a dataset (posted to bCourses), and identify a research question you could test using the variables in the dataset. Make sure that the DV is measured as a continuous variable; the IV can be either continuous or categorical. Note that your exam will require creating a scale for one of the variables.
Problem 2 : Load (and check) the Data and Report the Sample Size.
Problem 3 : Clean / Create a Scale and Describe the Variables You Will Use in the Model. Graph each variable that will be in your model. Make sure to do any scale creation needed, and remove any outliers / empty levels as necessary. Report descriptive statistics (mean, median, SD, and range) for the continuous variables. You should also report the alpha reliability for the scale you create.
Problem 4 : Define, Plot, and Interpret Linear Models to Test Your Theory. Define the linear model (DV ~ IV), plot this model, and report the intercept, slope(s), and R2 value from this model. How much do you think this pattern is due to sampling error (use bootstrapping OR NHST). What do you learn from this model? Was the researchers theory supported or not? What are the FOUR reasons this pattern might occur?
Problem5: Find one research article (for your final project, or otherwise). What was the main research question and result of this article? What is the population for this research question? What was the sample? Was this sample biased? How might this bias have influenced the results (if at all)?
