Week 6 Overview
| Dates | 24 February 2025 - 28 February 2025 |
| Reading | Required: SCIU4T4 Workbook chapters 20 and 21 |
| Recommended: Navarro and Foxcroft (2022) Chapter 11 | |
| Suggested: None | |
| Advanced: Johnson (1995) (Download) | |
| Lectures | 6.1: What is hypothesis testing? (11:18 min; Video) |
| 6.2: Making and using hypotheses (10:32 min; Video) | |
| 6.3: The example of the right-handed European toad (16:43 min; Video) | |
| 6.4: Errors, hypothesis testing and CI intervals (16:43 min; Video) | |
| 6.5: Student’s t-distribution and 1 sample t-test (13:31 min; Video) | |
| 6.6: Independent and paired samples t-tests (18:21 min; Video) | |
| 6.7: Handling violations of assumptions (7:50 min; Video) | |
| 6.8: Non-parametric tests (13:30 min; Video) | |
| Lecture | Lecture: 26 FEB 2025 (WED) 09:00-10:00 Cottrell LT B4 |
| Practical | Hypothesis testing and t-tests (Chapter 23) |
| Room: Cottrell 2A15 | |
| Group A: 26 FEB 2025 (WED) 10:00-13:00 | |
| Group B: 27 FEB 2025 (THU) 15:00-18:00 | |
| Help hours | Martina Quaggiotto |
| Room: Cottrell 2Y8 | |
| 28 FEB 2025 (FRI) 14:00-16:00 | |
| Assessments | Week 6 Practice quiz on Canvas |
Week 6 introduces hypothesis testing, and how to use and interpret statistical tests that test whether or not the mean (or median) of a dataset is significantly different from some specific value, or whether two different groups in a dataset have the same mean (or median).
Chapter 21 introduces hypothesis testing, and what it means in statistics. This chapter discusses the general idea of hypothesis testing using the concrete example of coin flipping to explain statistical terminology (null and alternative hypotheses, and p-values). The concepts and terminology in this chapter are essential for understanding the more advanced tests in later chapters. Examples of null and alternative hypotheses in the biological and environmental sciences are provided, and statistical significance is introduced as a concept along with the p-value. The chapter ends with an explanation of the difference between a Type I error and a Type II error, and the logic behind rejecting or not rejecting a null hypothesis.
Chapter 22 introduces the t-test and its non-parametric alternatives. Tests include the one sample t-test, the independent samples t-test, the paired sample t-test, the Wilcoxon test, and the Mann-Whitney U test. The chapter also explains the assumptions underlying these different tests. For each test, an example is provided of student test scores. For the one sample t-test, the null hypothesis that the mean of test score equals a specific value is tested against an alternative hypothesis that test scores are greater than the value. The t-statistic is calculated manually and its position on the x-axis of the t-distribution is illustrated to show how a p-value is obtained. The same pattern of explanation is followed with the independent samples t-test and the paired samples t-test, with test scores being used at the focal example for each test. Assumptions of continuous data, random sampling from the population, and sample means being normally distributed around the true mean are explained. The chapter then explains what to do if these assumptions are violated, with a brief example of log transforming positively skewed data followed by more in-depth explanation of the non-parametric alternative tests, the Wilcoxon test and the Mann-Whitney U test.
Chapter 23 is a guide for using jamovi to run the statistical tests that are introduced in Chapter 22. It includes five exercises, one for each of the Chapter 22 statistical tests. Exercises focus on a hypothetical example of biology education, specifically scores from biology students on different tests, and across two different years. Exercises focus on test scores from students and explain how to find and report key statistics in jamovi, including t-statistics, degrees of freedom, and p-values. These exercises also address when to reject the null hypothesis based on p-values, and how to interpret the outcomes of each test. How to apply non-parametric tests in jamovi is also introduced along with how to use jamovi to test t-test assumptions. Tests of model assumptions include a homogeneity test (Levene’s test) and a normality test (Shapiro-Wilk test), and the chapter also explains how to make inferences from a Q-Q plot.