Variables & Experimental Design
In an experiment, this is the variable that the researcher directly manipulates, controls, or alters
Independent Variable
This hypothesis asserts that the independent variable will have no significant effect on the dependent variable—a statement of "no difference"
Null Hypothesis
Designed to eliminate negative numbers and decimals found in z-scores, this standard score scale sets the mean at 50 and the standard deviation at 10.
T Score
(z-score of -1.0 equals a T score of 40)
Originating from Western Electric plant studies, this phenomenon occurs when subjects alter or improve their performance simply because they know they are being observed
Hawthorne Effect
This basic parametric test is used to compare the means of two sample groups to determine if they differ significantly
t-test
This variable represents the outcome measurement or data generated by the experiment
Dependent Variable
In counseling and social science research, this is the accepted default probability level or cutoff point for statistical significance
p</= .05
The highest level of measurement, this scale features equal intervals and an absolute, non-arbitrary true zero point
Ratio
Also known as the experimenter expectancy effect or Pygmalion effect, this threat occurs when a researcher's beliefs subtly influence subjects to fulfill expected outcomes
Rosenthal Effect
Developed by Ronald Fisher, this parametric statistic is used to compare the means of three or more groups with one independent variable
One Way ANOVA
Also known as Occam's Razor, this research principle dictates that experimenters should interpret findings using the simplest and most economical explanation
Parsimony
Also called an alpha error, this mistake occurs when a researcher rejects a true null hypothesis
Type I Error
In a normal bell-shaped distribution, this exact percentage of scores falls within +/- 1 standard deviation of the mean
68%
Named after a famous steel driver, this threat occurs when control group members engage in compensatory rivalry to outperform the treatment group
John Henry Effect
This nonparametric test serves as the distribution-free alternative to the independent t-test when comparing two uncorrelated groups
Mann-Whitney U test
This true experimental design utilizes two control groups and two experimental groups to detect and rule out pretest sensitization effects
Solomon Four-Group Design
This mistake occurs when a researcher accepts or fails to reject a false null hypothesis
Type II Error
On a standard frequency polygon graph, the horizontal x-axis used to plot the independent variable is called this
Abscissa
The tendency for unusually high or abnormally low scores on an initial test to move closer to the group average upon retesting
Statistical Regression to the Mean
This extension of the ANOVA statistically controls or adjusts for an extraneous confounding variable known as a covariate
ANCOVA
In single-subject N=1 research, this reversal design tracks baseline (A), treatment (B), withdrawal of treatment (A), and re-implementation of treatment (B)
ABAB Design
Expressed mathematically as 1-beta, this property describes a statistical test's capacity to correctly reject a false null hypothesis
Statistical Power
Describing the peakedness of a distribution, this term refers to a curve that is flatter and more spread out than a normal curve
Platykurtic
When deception is unavoidable in an experiment, ACA and NBCC ethical guidelines dictate that researchers must perform this procedure with subjects as soon as possible following the study
Debriefing
The most widely used nonparametric statistical test, this measure determines whether observed frequency data differ significantly from expected theoretical frequencies across mutually exclusive nominal categories
Chi Square