Statistics in Psychology

Study Notes

Statistics transforms data into information and practice, crucial for an information-based society.
Understanding basic statistics is essential to avoid manipulation and ensure informed decision-making.
Statistics is a critical mathematical discipline for processing data into meaningful information.
Basic statistical knowledge empowers individuals to interpret data encountered in everyday life.
Psychological conclusions about human behavior need statistical methods to derive valid conclusions.
Data collection requires sampling from populations to minimize bias.
Rigorous statistical analysis supports hypothesis testing.
This chapter provides a refresher course on statistics and its use in psychological assessment.

Statistics concisely describes observations, comparing them to one another.
Logical conclusions about unobserved events can be drawn using statistics.
Descriptive statistics describes data in a concise form
Inferential statistics uses data from a segment to draw conclusions about a larger group

Scales assign numbers to characteristics following rules.
Continuous scales measure continuous variables.
Discrete scales categorize variables without numerical meaning.
Scale properties include magnitude (more/less), equal intervals (same meaning using scale units), and absolute zero (absence of the characteristic.)
Examples of psychological tests (IQ levels) rarely exhibit equal intervals or absolute zero points.

Nominal scale: categorizes objects.
Ordinal scale: ranks objects but does not define differences between ranks.
Interval scale: ranks objects with equal interval sizes but no absolute zero.
Ratio scale: ranks objects with equal interval sizes and an absolute zero point.

Frequency distributions: show how often each score occurs.
Graphic forms: represent data visually with Histograms, Bar Graphs, and Frequency Polygons.

The normal curve displays a bell-shaped distribution.
Mean, median, and mode are all located at the peak.
Standard deviation units measure distance from mean in either direction.
A well-established theory describes the distribution patterns under ideal conditions and is useful for inferential testing.

Correlation measures the relationships between variables.
Correlation coefficients (e.g., Pearson r, Spearman ρ) quantify the strength and direction of these relationships.
A correlation does not imply causation. (Correlation does not equal causation).
Positive, negative, and no correlation are indicative of the relationship between two variables.

Pearson r is commonly used to examine linear relationships between continuous variables.
Spearman ρ measures the relationship between ranked data.
Point-biserial correlation measures association when one variable is dichotomous (e.g., yes/no) and the other is continuous.
Phi coefficient is used for two dichotomous variables.