Intro to Statistic Analysis

Written By Ariel Tan

This lesson plan was a part of the "EDUC 198 Directed Research in Education course" which I started teaching in Fall 2021. Course participants are Research Assistants (RAs) for the UCI Working Memory and Plasticity Lab (WMP Lab). This course is an introductory and supplementary class for undergraduate/new RAs who are doing research for course credit. In addition to their lab responsibilities, RAs are required to complete the assigned readings and a final project.

Introduction to Statistics and Statistical Program

For this week, RAs are watching a brief video that provides an overview of statistics. The main takeaway from this video (along with this course series) is to understand that statistics are tools that provide insights rather than answers to our research questions. 

What Is Statistics: Crash Course Statistics #1Links to an external site.

For the course assignments, we are instructing RAs to perform basic data cleaning and analysis via JASP and Google sheet.

Cleaning Raw Data

Data cleaning is a very important but overlooked foundation for statistics. When people hear the word "analysis", they often think of something much more complicated, advanced, and math-orientated.  

However, statistics also has an admin-heavy side that is critical to your results. Assuming that you have collected data properly and ethnically, data cleaning is the first step to analysis that facilitates later analysis.

For this week, we are asking RAs to read the following blog posts that provide brief, simple, but good tips and instructions on the basics of data cleaning. Please note that most mature researchers clean data in their preferred software (or have their preferred method), but for the purpose of this course, we will start with cleaning data using Google Sheets.

Importance of Good Data Collection and Descriptive Analysis

A cleaned data spreadsheet allows you to perform simple to complex analysis, but before then, we need to understand what is the story and context that our data is trying to provide. For example: What does our sample look like? What can we learn about the general story (or "vibe" if you may) of this study?

The first reading for this week is a chapter from a Department of Education report on descriptive analysis in education. It describes ways that descriptive analysis can provide helpful information for studies - and that sometimes, a descriptive analysis itself can even be a standalone research result. 

The second reading for this week features a PNAS essay that talks about common errors in data collection. While the whole paper is very relevant and informative, you can focus on sections entitled “Why Focusing on Errors Is Important” and “Underlying Themes of Errors and Their Contributing Factors” for now. 

  • Brown, A. W., Kaiser, K. A., & Allison, D. B. (2018). Issues with data and analyses: Errors, underlying themes, and potential solutions. Proceedings of the National Academy of Sciences of the United States of America, 115(11), 2563–2570. https://doi.org/10.1073/pnas.1708279115

Additionally, we included an optional reading about skewness and Kurtosis in assessing normal distribution for those who want to further advance their understanding of distributions. When you collect a group of data and before you can choose an analysis method, there is a set of parameters you need to meet that guides you to select the appropriate tests. One of the command parameters is that the sample data needs to be normally distributed, which requires us to test the distribution using skewness and Kurtosis.

  • Kim H. Y. (2013). Statistical notes for clinical researchers: assessing normal distribution (2) using skewness and kurtosis. Restorative dentistry & endodontics, 38(1), 52–54. https://doi.org/10.5395/rde.2013.38.1.52

Hypothesis Testing

For this week, we will be learning about inferential statistics that allow us to draw results and conclusions from the data. It's also the calculation and interpretation that people often think of when they hear the words "analysis". To do so, we will start with learning about probability theory and hypothesis testing.

T-Test

For this week, we are learning one of the most common analyses that compare group means.

Correlations

Correlation is a term that's commonly used in daily language, and even non-scientists are familiar with the saying "correlation does not equal causation". For this week's reading, we have a journal essay that talks about the proper usage of correlation. Though the audience is geared towards the medical field, the content is still applicable in psychology and education (and any type of science actually).

Ariel Tan
Previous

Gap Year: 如何提高学术竞争力
Next

Professional Exploration