Page 1 of 8
COMP517 – Data Analysis
Lab 9 – Introduction to One-way ANOVA and Simple Linear Regression
Introduction
In this lab, we introduce the fundamental concepts of hypothesis testing using one-way analysis of
variance (ANOVA) and simple linear regression models. We will use ANOVA statistical techniques to
analyze exam scores across various streams of students. Then we perform Tukey's Honestly Significant
Difference (HSD) post hoc test after conducting an analysis of variance (ANOVA) to compare test
scores among different streams. In the latter part of the lab, we will employ the identical dataset for
conducting a linear regression analysis. Simple linear regression serves various purposes, but in this
lab, we concentrate on specific objectives. It enables the prediction of the dependent variable by
leveraging the independent variable, facilitating the estimation, or forecasting of outcomes for new or
unobserved data points. In addition, it provides a valuable means of understanding relationships
between variables. The slope coefficient (β1) within the regression model quantifies how variations in
the independent variable influence changes in the dependent variable while holding all other factors
constant, thus offering insights into the cause-and-effect dynamics between the variables of interest.
A preliminary version of this lab is made available under the name "Lab 09_Preload.ipynb". This
Python file comes with explanatory comments and some existing codes designed to guide you through
the tasks. Download the file and save it your lab 09 directory.
Lab Objectives
✓ One-Way ANOVA: One-way analysis of variance (ANOVA) is a statistical method to
compare the means of three or more groups. We will extend our analysis to multiple streams
and assess whether there is a significant variance in exam scores across these streams.
✓ Tukey's Honestly Significant Difference (HSD) Test: Conduct Tukey's HSD post hoc test
following an analysis of variance (ANOVA) to assess and compare test scores across various
stream categories.
✓ Simple Linear Regression: One of the objectives of simple linear regression is to establish and
understand the relationship between two variables: one is the dependent variable (the variable
you want to predict or explain), and the other is the independent variable (the variable used to
make predictions or provide explanations). This analysis then aims to predict a student's test
score based on their performance in the lab test. Following that, we will provide a concise
discussion on the findings of the model.
Data Source
Our dataset comprises exam scores from students in different streams of a given course. By applying
these hypothesis testing techniques, we aim to uncover potential disparities in academic performance
among these streams. Download the ‘student_streamscores.csv’ and save it under your lab 09 directory.
Assumptions
For this lab, we assume that our data follows a normal distribution, and we also assume that the key
assumptions for linear regression validity are met. We will cover these assumptions next week.
资源评论
