Welcome to Validity, Reliability, and Data Analysis!

Ever wondered why some psychology studies are "better" than others? Or why we can't always trust a single experiment? It usually comes down to how valid or reliable the research is. In this chapter, we will learn how to judge a study’s quality and how psychologists make sense of the mountains of data they collect. Don’t worry—we’ll break the math down into simple steps!

1. Validity: Is it True?

Validity refers to the accuracy of a study. It asks: "Is the researcher actually measuring what they claim to be measuring?"

Types of Validity

  • Ecological Validity: This is about how well the findings apply to the real world. If a study takes place in a very artificial laboratory (like Andrade’s doodling study), it might have lower ecological validity than a study in a natural setting (like Piliavin’s subway study).
  • Temporal Validity (A Level only): This asks if the findings are still relevant today. For example, does a study on social behavior from 1950 still apply in 2024, or have cultural changes made the results "out of date"?
  • Generalisability: Can the results from the small group of participants (the sample) be applied to the wider population? If a study only uses university students, it might not be generalisable to elderly people.

Threats to Validity

  • Demand Characteristics: This happens when participants guess the aim of the study and change their behavior to "help" or "hinder" the researcher. This makes the results less valid because the behavior isn't natural.
  • Subjectivity: When a researcher's own personal feelings or opinions influence the results. Objectivity is the opposite—it's when the data is factual and not influenced by the researcher’s bias.

Quick Tip: Think of validity like a bullseye. If you hit the center of the target, you are being valid (accurate). If you miss the target entirely, your measurement is "off."

2. Reliability: Is it Consistent?

Reliability is all about consistency. If you did the study again, would you get the same results?

Key Reliability Terms

  • Replicability: This is the ability for a study to be repeated exactly. To be replicable, the researchers must use standardisation—keeping everything the same for every participant (like using the same instructions or the same room).
  • Test-retest Reliability: This involves giving the same test to the same person at two different times. If the scores are similar, the test is reliable.
  • Inter-rater Reliability: This is used when researchers are observing behavior. It’s when two or more observers watch the same thing and agree on what they saw. If they have high agreement, the data is reliable.

Common Mistake to Avoid: Many students confuse validity and reliability.
Reliability = Consistency (Like a scale that always says you weigh 5kg more than you do—it's consistent, but wrong!).
Validity = Accuracy (Like a scale that shows your true weight).

3. Types of Data

Psychologists collect two main types of data. Most studies use a mix of both!

  • Quantitative Data: Numerical data (numbers).
    Example: A score on a memory test or the time taken to complete a task.
    Strength: Easy to compare and analyze using statistics.
  • Qualitative Data: Descriptive data (words).
    Example: A participant’s description of how they felt during an experiment.
    Strength: Provides rich, deep detail about why someone behaved a certain way.

Did you know? At A Level, you will also learn about Psychometric Tests. These are standardized tests designed to measure mental characteristics, like personality or intelligence, in a quantitative way.

4. Data Analysis: Descriptive Statistics

Once we have our data, we need to summarize it. We use Measures of Central Tendency to find the "middle" and Measures of Spread to see how varied the scores are.

Measures of Central Tendency

You don't need to do complex math for the exam, but you must be able to recognize and find these:

  • Mode: The most frequent score in a data set.
  • Median: The middle score when all numbers are put in order.
  • Mean: The average score (add all scores together and divide by the number of scores).

Measures of Spread

  • Range: The difference between the highest and lowest score.
    \(Range = Highest\ value - Lowest\ value\)
  • Standard Deviation (SD): This tells us how much the scores vary from the mean.
    A low SD means most scores are close to the average (the group is similar).
    A high SD means the scores are spread out (the group is very different).

5. Visualising Data: Graphs

Sometimes a picture is worth a thousand numbers! You should know which graph to use for different types of data.

  • Bar Charts: Used for nominal data (categories). There are gaps between the bars.
    Example: Comparing the mean aggression scores of boys vs. girls.
  • Histograms: Used for continuous data. The bars touch each other.
    Example: Showing the distribution of ages in a study.
  • Scatter Graphs: Used for correlations. Each dot represents a participant’s score on two different variables.
    Example: Comparing hours of sleep with test scores.

Key Takeaway Summary

Validity = Accuracy (Does it measure what it should?).
Reliability = Consistency (Can we repeat it and get the same result?).
Quantitative = Numbers; Qualitative = Words.
Mean, Median, Mode find the middle; Range and SD show the spread.
Standardisation is the "glue" that makes a study replicable!

Don't worry if standard deviation seems confusing at first—just remember it's a way of seeing if your participants were all similar or very different from each other!