Lesson 10.4: Running and Reporting a Statistical Analysis

Introduction

Welcome, students! In this lesson, we will dive into an essential skill in psychology: running and reporting a statistical analysis. This unit will help you become more comfortable with numbers and data interpretation, which are crucial in the field of psychology. 🌟

Learning Objectives

By the end of this lesson, you should be able to:

Work through a simple non-parametric test using a concrete example.
Compare a calculated statistic to a critical value and state the outcome.
Accept or reject the null hypothesis and articulate your conclusion.
Report statistical results in the standard format used in research.
Use software tools for running basic statistical analyses.

Understanding Non-Parametric Tests

Non-parametric tests are statistical tests that do not assume a specific distribution of the data. They are particularly useful when you have small sample sizes or when the data you are working with does not meet the assumptions needed for parametric tests like t-tests. In psychology, we often deal with ordinal data or non-normally distributed interval data, which makes non-parametric tests very useful.

Example: The Mann-Whitney U Test

Let’s explore an example to better understand how to execute a non-parametric test, specifically the Mann-Whitney U test, which compares the ranks of two independent groups. Suppose we want to understand if two different teaching methods impact student scores differently.

Step 1: Gather Data

Let’s say we have the following scores from two teaching methods:

Method A: [78, 85, 88, 90, 70]

Method B: [65, 70, 74, 80, 66]

Step 2: Rank the Data

We first combine the two groups and rank the scores. Here’s how the ranks would look:

Combined scores: [65, 66, 70, 74, 78, 80, 85, 88, 90]
Ranks: [1, 2, 3, 4, 5, 6, 7, 8, 9]

Where each score is given a rank based on its value. The lowest score receives a rank of 1.

Method A Ranks: [5, 7, 8, 9, 3]
Method B Ranks: [1, 2, 4, 6, 3]

Step 3: Calculate U Statistic

Now, we can calculate the U statistic for both groups using the formula:

$$U = R - \frac{n(n + 1)}{2}$$

Where:

$U$ is the U statistic,
$R$ is the sum of ranks for one of the groups,
$n$ is the number of observations in that group.

For Method A:

Sum of ranks ($R_A$) = 5 + 7 + 8 + 9 + 3 = 32
Number of observations ($n_A$) = 5
So, $U_A = 32 - \frac{5(5 + 1)}{2} = 32 - 15 = 17$

For Method B:

Sum of ranks ($R_B$) = 1 + 2 + 4 + 6 + 3 = 16
Number of observations ($n_B$) = 5
So, $U_B = 16 - \frac{5(5 + 1)}{2} = 16 - 15 = 1$

Step 4: Compare U to Critical Value

Next, we’ll compare our calculated U values with a critical value from the Mann-Whitney U distribution table. Let’s assume the critical value at significance level $\alpha = 0.05$ for $n_A = n_B = 5$ is 2.

Since $U_B = 1 < 2$, we reject the null hypothesis. This indicates significant differences between the two teaching methods regarding their effectiveness.

Reporting the Results

Now, let’s talk about how to clearly report our findings. In psychology, it's important to present your results in a specific format. Here’s how we could report our findings from the Mann-Whitney U test:

“A Mann-Whitney U test was conducted to determine if there was a difference in scores between students taught by Method A (Mdn = 85) and Method B (Mdn = 70). The results indicated a significant difference in scores, U = 1, p < 0.05, with scores from Method A being higher than those from Method B.”

Using Software for Statistical Analysis

Computers are fantastic tools that can help us run statistical tests more efficiently. Software like Excel or statistical packages like SPSS can make the process seamless.

For instance, if you were using Excel to perform the Mann-Whitney U test, you would input your data into different columns, select the appropriate functions or add-ins for analysis, and run your calculations without hand-ranking your data manually. 🚀

Conclusion

In this lesson, students, we have covered the basics of running and reporting a statistical analysis focusing on non-parametric tests like the Mann-Whitney U test. Understanding how to interpret and report statistical tests is essential in psychology, as it informs research conclusions and contributes to a body of knowledge.

Getting comfortable with both the manual calculations and software tools allows you to access a realm of quantitative analysis that is indispensable in our field.

Study Notes

Non-parametric tests do not assume a specific data distribution.
The Mann-Whitney U test compares scores between two independent groups.
To calculate U: $$U = R - \frac{n(n + 1)}{2}$$
Report results in the format: method, median, U statistic, and p-value.
Software can simplify complex calculations and data analysis.