Section 7.11 Conclusion
In this chapter, we examined how studying time, exam anxiety, and exam performance are related to one another using correlation-based techniques. Our analyses revealed several important patterns:
-
Students who studied more tended to earn higher exam scores.
-
Higher anxiety levels were associated with lower exam performance.
-
Students who studied more also tended to report lower anxiety before the exam.
However, these relationships did not exist in isolation. When we controlled for studying time using a partial correlation, the relationship between anxiety and exam performance weakened substantially, and the relationship between studying time and exam performance was no longer statistically significant after controlling for anxiety. This highlights an important statistical insight: relationships between variables often overlap, and failing to account for this overlap can lead to misleading interpretations.
Throughout this chapter, we also reinforced a critical principle in data analysis:
Correlation does not imply causation.
While studying, anxiety, and performance are clearly related, correlation alone cannot tell us whether studying causes better performance, whether anxiety reduces scores, or whether some third factor influences all three. Correlation helps us describe relationships — not explain them.
By visualizing relationships, computing correlation coefficients, interpreting r and r², and applying partial and point-biserial correlations, you now have a toolkit for understanding how variables move together and how those relationships change when additional factors are considered.
In the next chapter, we will build on this foundation by moving beyond description and toward prediction. Using what we’ve learned about relationships between variables, we’ll begin constructing regression models that allow us to estimate outcomes — not just observe associations.
