Clever Grades

🎧 Read Aloud

Correlational Analysis Principles

Defining Correlation

What is it?

Correlation is a statistical technique used to measure and describe the strength and direction of the relationship between two variables. Correlations do not involve manipulation and cannot determine causation but can show whether variables are associated.

Obtaining Data

The strength of a correlational study relies heavily on the quality and nature of the data collected. Here is the typical process:

1

Measurement

Data for correlation usually comes from measuring two continuous variables for each participant or case.
2

Sources

Measures may come from observations, self-reports, or secondary data sources.
3

Paired Scores

After data collection, each case has a pair of scores — one for each variable.
4

Archival Use

Sometimes data is gathered from archival sources (e.g., crime rates vs. unemployment).

Correlation Coefficients

A correlation coefficient quantifies the degree and direction of the relationship between variables. The value typically ranges between -1.00 and +1.00:

➕

+1.00

Perfect positive correlation.
➖

-1.00

Perfect negative correlation.
0

Zero

No correlation.
🧮

Pearson’s r

The most commonly used coefficient, measures linear relationships between variables measured on interval or ratio scales. For ordinal data, Spearman’s rho may be used.

Direction of Relationship

⬆️
Positive Correlation When both variables increase or decrease together. For example, as study hours increase, exam scores increase. The data points trend upwards on a scatter diagram.
⬇️
Negative Correlation When one variable increases as the other decreases. For example, as stress levels increase, happiness scores decrease. Data points trend downwards on a scatter diagram.

Visualizing the Data

No Correlation

No apparent relationship between two variables. Data points scattered randomly, no clear pattern. For example, shoe size and intelligence would typically show no correlation.

Scatter Diagrams

Visual representation of the relationship between two variables. Each point represents one participant’s paired scores. Patterns indicate the type and strength of correlation.

The Crucial Caveat

⚠️

Correlation does not imply causation: The presence of a strong relationship does not mean one variable causes the other. This is the primary limitation of correlational studies.

Other Limitations

Correlations can be tricky. Here are some factors that can distort or complicate the interpretation of results:

🤔
What are common pitfalls in correlational analysis?
🦉
Correlations may be influenced by outliers, restricting range, or third variables.
🤔
What is the directionality issue?
🦉
It may be unclear which variable influences the other (Directionality problem).
```
Correlation Deck
Q
What does correlation measure?

What does correlation measure?

A
Answer

The strength and direction of the relationship between two variables.

Q
Can correlation determine causation?

Can correlation determine causation?

A
Answer

No, correlation cannot establish causation.

Q
What is the range of a correlation coefficient?

What is the range of a correlation coefficient?

A
Answer

From -1.00 to +1.00.

Q
What does a correlation coefficient of +1.00 represent?

What does a correlation coefficient of +1.00 represent?

A
Answer

A perfect positive correlation.

Q
What does a correlation coefficient of -1.00 indicate?

What does a correlation coefficient of -1.00 indicate?

A
Answer

A perfect negative correlation.

Q
What is Pearson’s r used for?

What is Pearson’s r used for?

A
Answer

Measuring linear relationships between interval or ratio variables.

Q
What correlation measures are used for ordinal data?

What correlation measures are used for ordinal data?

A
Answer

Spearman’s rho or other non-parametric correlations.

Q
What does a positive correlation signify?

What does a positive correlation signify?

A
Answer

Both variables increase or decrease together.

Q
What is an example of negative correlation?

What is an example of negative correlation?

A
Answer

Stress levels increase while happiness scores decrease.

Q
What does no correlation look like on a scatter diagram?

What does no correlation look like on a scatter diagram?

A
Answer

Data points are scattered randomly with no clear pattern.

Q
What is a scatter diagram?

What is a scatter diagram?

A
Answer

A visual representation of paired scores for two variables.

Q
Name one limitation of correlation analysis.

Name one limitation of correlation analysis.

A
Answer

It cannot confirm causation.

Q
What problem arises from correlation regarding variable influence?

What problem arises from correlation regarding variable influence?

A
Answer

Directionality problem—uncertainty about which variable influences the other.

Q
Where can data for correlation come from?

Where can data for correlation come from?

A
Answer

Observations, self-reports, secondary or archival sources.

Tip: Tap a card to flip. Shuffle or Reset anytime.

📈 Correlation Quiz

1. Which of the following correlation coefficients indicates the strongest positive relationship?

Correlation coefficients closer to +1 indicate stronger positive relationships.

2. What does a correlation coefficient of 0 mean?

Zero correlation means there is no observable relationship between variables.

3. Which statistical method is most appropriate for correlation between two ordinal variables?

Spearman’s rho is used to measure correlation with ordinal data.

4. Why can correlation not establish causation?

Correlation only shows association, not which variable causes the other.

5. If a scatter diagram shows data points trending downwards, what type of correlation is indicated?

A downward trend suggests one variable increases while the other decreases.

📊 Results

  • Notes and Quiz
  • Psychology
  • AS Level
  • Research Methods
  • OCR