Much to Be Learned

I was enrolled in a research methods class at Duke with Alison Koenka this summer and learned so much about measurement, validity, reliability, and the different between-group and within-group research designs commonly used in psychology and neuroscience. Thus, I feel very comfortable with experiments. However, despite having taken a statistics course already at Duke, I am still confused by statistical methods and find it difficult to make sense of them. At a basic level, I understand that there are descriptive and inferential statistics, and both are necessary for helping researchers organize, summarize, and interpret the data in a way that some sense can be made from it. I also remember the theory behind null hypotheses testing and do find the alpha significance level of a comparison distribution to be important in separating research that reports true effects in a population from research that does not. However, though I am familiar with the normal distribution depicted below, I find myself lost in the details, especially with regard to the statistical theory behind the F and t distributions. In addition, while I do know about type I and type II errors, I do not understand what ‘power’ is and how some evidence has greater power than others. 

Despite my previous knowledge of research design and statistics, I still find journal articles confusing as well. Many times, articles use methods with which I am not familiar. For this reason, I am confused by the graphs and tables provided describing the experiment’s data. For example, I still do not understand representations indicating EEG or TMS data. However, it is my hope that greater exposure to methods used by researchers in psychology and neuroscience will help me understand these graphs and images better.