Reporting and Presenting T-test Conclusions

TL;DR

After running a t-test, you need to clearly and concisely communicate your findings, including the specific t-test used, its results, and what they mean. The goal is to inform your audience whether there's a statistically significant difference between group means. Always include essential statistics like t-value, degrees of freedom, and p-value.

1. The Mental Model

Think of reporting your t-test like telling a clear, evidence-based story. You introduce the question, present the statistical evidence (the t-test results), and then deliver the punchline: what did you learn about the groups you compared?

2. The Core Material

Once you've run a t-test and have your results, the next crucial step is to present them effectively. This means detailing what you did, what you found, and what it implies, all in a way that's easy to understand and scientifically sound.

Components of a Good Report

Photo by RDNE Stock project on Pexels

Every good t-test report should include:
1. What you did: State the type of t-test you performed (e.g., independent samples, paired samples, one-sample).
2. Why you did it: Briefly state the hypothesis or research question you were investigating.
3. Key statistics: Provide the t-statistic (t), degrees of freedom (df), and the p-value (p).
4. Means and standard deviations: Report the descriptive statistics (mean, standard deviation) for each group involved. This gives context to the difference you're discussing.
5. Conclusion: Clearly state whether you reject or fail to reject the null hypothesis, and what this means in terms of your original research question.

Reporting the p-value

Photo by Kindel Media on Pexels

The p-value is critical. It tells you the probability of observing your data (or more extreme data) if the null hypothesis were true.
* If p < 0.05 (or your chosen alpha level), the result is typically considered statistically significant. This means you reject the null hypothesis.
* If p ≥ 0.05, the result is typically considered not statistically significant. This means you fail to reject the null hypothesis.

When reporting the p-value, follow these guidelines:
* If p < .001, report as "p < .001".
* For other significant values, report the exact p-value (e.g., "p = .023").
* For non-significant values, report the exact p-value (e.g., "p = .156").

Standard Format for T-test Results

Photo by adrian vieriu on Pexels

A common way to present t-test results is:
"t(df) = t-value, p = p-value"

For example:
* "t(28) = 2.10, p = .045"
* "t(15) = -0.85, p = .409"

graph TD
    A["Identify Research Question/Hypothesis"] --> B["Choose Appropriate T-test"]
    B --> C["Perform T-test (e.g., in R, Python, Excel)"]
    C --> D{"Obtain Key Statistics:
    t-value, df, p-value"}
    D --> E["Extract Group Means and SDs"]
    E --> F{"Interpret p-value
    (p < α vs. p ≥ α)"}
    F --> G["Formulate Conclusion
    (Reject/Fail to Reject Null)"]
    G --> H["Write Clear Report
    (Include All Components)"]

Writing the Narrative

Photo by KoolShooters on Pexels

Beyond the numbers, you need to explain what they mean in plain language relevant to your study.

Example structure for an independent samples t-test:
"An independent samples t-test was conducted to compare [dependent variable] in [group 1] and [group 2]. There was a significant/no significant difference in [dependent variable] scores for [group 1] (M = [mean 1], SD = [SD 1]) and [group 2] (M = [mean 2], SD = [SD 2]); t([df]) = [t-value], p = [p-value]."

3. Worked Example

Let's say you conducted an independent samples t-test to compare the average test scores of students who used a new study method versus those who used a traditional method.

Hypothesis: Students using the new study method will have significantly higher test scores than those using the traditional method.

Results from your statistical software:
* New Method Group: Mean = 85, SD = 5, n = 30
* Traditional Method Group: Mean = 80, SD = 6, n = 32
* T-statistic: t = 3.50
* Degrees of Freedom: df = 60
* P-value: p = 0.001

How to report it:

"An independent samples t-test was conducted to compare the mean test scores of students using a new study method (n=30) versus those using a traditional method (n=32). Students who used the new study method had significantly higher test scores (M = 85.0, SD = 5.0) compared to students who used the traditional method (M = 80.0, SD = 6.0), t(60) = 3.50, p = .001. These results suggest that the new study method is associated with improved test performance."

4. Key Takeaways

• Always specify the type of t-test performed.
• Include the t-statistic, degrees of freedom, and p-value in a standard format: t(df) = t-value, p = p-value.
• Report descriptive statistics (means and standard deviations) for each group to provide context.
• Clearly state whether the difference was statistically significant and what that means for your hypothesis.
• Use clear, non-technical language when interpreting your findings for your audience.
• Be precise with p-values; report exact values or "p < .001" as appropriate.
• Remember that statistical significance doesn't always equal practical significance.

Common mistakes to avoid:
- Forgetting descriptive statistics: Just reporting t, df, and p isn't enough; context is key.
- Misinterpreting the p-value: A p-value doesn't tell you the strength of the effect, only the probability of observing data under the null hypothesis.
- Omitting the t-test type: Your audience needs to know which test you ran.
- Overstating conclusions: Avoid making causal claims if your study design doesn't support them (e.g., from observational data).
- Inconsistent formatting: Stick to a standard reporting format for clarity.

5. Now Try It

Imagine you ran a paired samples t-test to see if a weight loss program had an effect. Participants' weights were measured before and after the program.
* Before Program: Mean = 75 kg, SD = 8 kg
* After Program: Mean = 72 kg, SD = 7 kg
* T-statistic: t = 2.80
* Degrees of Freedom: df = 19 (for 20 participants)
* P-value: p = 0.011

Your task is to write a concise paragraph reporting these results, similar to the worked example. What success looks like: Your paragraph includes the test type, the means and SDs, the t-statistic, df, p-value, and a clear interpretive sentence about the program's effect.