(Source: Sommer, B and Sommer, R. A (1997). A Practical Guide to Behavioral Research)



Alternative Hypothesis: Working or research hypothesis, and alternative to the null hypothesis, stated in positive terms (i.e., that there will be an effect)


Analysis of Covariance (ANCOVA): Takes into account scores on some other variable believed to affect the action of the independent variable


Analysis of Variance (ANOVA): Statistical technique for comparing differences between means


Anonymity: Researcher does not know the identity of the participants in the study


Applied Research: Systematic procedures to answer pressing questions


Attitude Scale: Type of questionnaire designed to produce scores indicating the overall degree of favorability of a person’s attitude on a topic


Average: Generic term referring to various measures of central tendency- the mean, median, and mode


Bar Graph: A graphic representation of results using bars to indicate frequency of response


Blind Testing: In a sensory evaluation or drug research, blind testing indicates the subject is not aware of the stimulus.


Browser Program: Required for navigating the World Wide Web


Case Study: In-depth investigation of a single instance


Categorical: Refers to variables that have levels that are mutually exclusive


Causal Patterns: Relationships of cause and effect


Central Tendency: Number (statistic) that best summarizes the characteristics of a sample as a whole


Chi-Square (X Squared): Statistical test used with categorical data to test whether an obtained distribution of scores differs reliably from what would be expected by chance.


Coding: Transformation or reduction or raw data into a set of standard categories for statistical analysis


Comparative Rating Scales: Respondent compares individual with others in the same category


Concurrent Validity: Correlation of a test with present behavior, or with other existing tests or measures; one type of criterion validity


Confidentiality: Participant or respondent identity is known to researcher, but is not publicly revealed.


Confounding: Confusion of the effects of variables, resulting in the inability to determine which variable is the cause of the observed effect.


Constants: Qualities that do not vary


Construct Validity: Linkage of the test measurement to specific theoretical constructs, the relationship of a test to theory


Content Validity: Degree to which test items assess the domain that a test claims to cover, the relevance of the items to the behavior measured


Contingency Table: Table used to record the relationship between two or more variables. The observed frequencies are placed into the cells of the table.


Continuous Variable: Variable whose levels can take on any value within the lowest and uppermost limit of the variable, for example time or income


Control Group: Subject group that resembles the experimental group in every respect except that it is not exposed to the independent variable


Convenience sampling: taking what is available as your sample


Correlated measures: Groups being compared have been matched on some characteristic, or the same people have been tested before and after an experimental treatment


Correlation: Association between two sets of scores. Often expressed in terms of a correlation coefficient


Correlation Coefficient: Indicates the degree of relationship between two sets of scores, a number that can range from +1.0 to -1.0.


Criterion Validity: Relationship of the test scores to other measures of the same characteristics


Degrees of Freedom: (df): Number of values that are free to vary after certain restrictions have been placed on the data. Used to evaluate the results of various statistical tests; related to the sample size and number of levels of the independent variable


Dependent Variable: Consequence or outcome or the manipulation, the variable that is affected by the independent variable; synonyms are outcome, response, or criterion variable


Descriptive statistics: Techniques to organize and summarize data


Equivalent forms: Two different, but comparable, forms of a scale; used for assessing reliability or to avoid practice effects


ERIC: Educational Resources Information Center of the U.S. Department of Education; also an electronic and print database


Expected Frequencies (E): Represent the null hypothesis- what you would expect by chance


Experimental Group: Subjects exposed to the levels of the independent variable; also called the treatment group


Experimental Variable: Variable that is manipulated or systematically altered by the experimenter; synonyms are independent variable or predictor variable


External Validity: Generalizability of research findings


Extraneous Variables: Variables, in addition to the independent variable, that might be affecting the dependent variable (outcome)


Face Validity: Appearance of being a valid measure of something; a measure that “looks right” to an outside observer. A form of content validity


Factor: Independent variable or treatment. A study investigating the effects of age and education on attitude would be a two-factor study.


Factorial Design: ANOVA design involving more than one independent variable (factor)


Focus Group: Group interview designed to explore what a specific set of  people think and feel about a topic


Format: Structure; pertains to the appearance, order, and wording of items in a questionnaire or interview; in contrast to content, which refers to meaning


Frequency: Number of times a score or category level occurs


Frequency Distribution: Arrangement of scores from highest to lowest, together with the frequency of each score


Histogram: Particular type of bar graph


Hypothesis: testable statement logically derived from theory or observation; can be either confirmed (accepted) or disconfirmed (rejected)


Independent Variable: Variable that is manipulated in order to measure its effect on some outcome; synonyms are experimental variable, or predictor variable


Inferential statistics: Statistical techniques used to make generalizations from samples to populations


Informed Consent: Those who participate in research studies should understand what is involved and freely consent to participate


Institutional Review Board (IRB): Group of people with formal responsibility for reviewing submitted research proposals in terms of ethics and protection of the participants


Interaction: Using ANOVA, the outcome produced by changes in one variable differs depending upon the levels of a second variable


Internal Validity: Degree to which a procedure measures what it is supposed to measure


Interval Scale: Level of measurement that provides information about size or direction, plus having equal intervals between scale points


Item Analysis: Shows the degree that the various items in a scale or test “hang together”


Likert Scale: Type of attitude scale containing statements that are clearly favorable or clearly unfavorable, to which respondent indicates degree of agreement


Matched Groups: Assigning subjects so that the experimental and control groups are as similar as possible.


Mean Comparisons: Comparisons or contrasts between individual means to locate where the significant difference lies.


Median: Midpoint of a distribution when all the scores are arranged from highest to lowest.


Mode: Single score that occurs most often in a distribution.


Nominal Measures: Characteristics assigned to categories. No underlying continuous dimension


Normal Curve: Symmetrical bell-shaped curve which often approximates the frequency of occurrence of events in nature


Null Hypothesis: Assumes that differences produced by the research manipulation are due to chance fluctuations and that the independent variable has no effect on the dependent variable. Generally, the researcher hopes to disprove the null  hypothesis.


Ordinal Scale: Characteristics can be ordered along an underlying dimension, but no information is provided about the distance between points; only provides information about increasing or decreasing size or direction


Pilot Study: Preliminary use of a procedure designed to identify problems and omissions before the actual study is conducted


Positive correlation: An increase in one variable is accompanied by an increase in the other


Probability Level: When used in statistics, it indicates the likelihood that an obtained difference on a statistical test is due to chance alone.


Probability Sample: Sample that is drawn in such a way that the probability for the inclusion of nay given individual can be estimated. Two general types of probability samples are random samples and stratified samples


Purposive Sample: Type of non-probability sample in which the individuals considered most relevant to the issue studied are selected for inclusion


Quasi-experiment: Non-random assignment of subjects to conditions; the experimenter lacks direct control over the independent variable; also called a natural experiment


Random sample: Type of probability sample in which every individual in the entire population being studied has an equal chance of being selected


Range: Measure of dispersion or variability, computed by subtracting the lowest from the highest score


Ratio Scale: Level of measurement that contains information on direction, possesses equal intervals, and an absolute zero point


Regression: Pertains to the correlation; the relationships between sets of scores


Reliability: Consistency in measurement; the repeatability or replicability of findings; stability of measurement over time


Reliability Coefficient: Coefficient of correlation between two administrations of a test or other measures


Repeated Measures Design: Research design involving the same subjects studied at different times or under different conditions


Sample: Subset of the population


Sample Bias: Error introduced by a sampling procedure that favors certain characteristics over others


Sampling Error: Chance variation among samples selected from a single population


Snowball Sample: Type of purposive (non-random) sample in which the researcher asks respondents for other people that should be contacted


Split-half Reliability: Dividing a scale or test into two halves which are then compared


Standard Deviation: Measure of dispersion or variability; abbreviated as s or SD

Stratified Sample: Type of probability sampling in which the characteristics of the sample are selected to be proportionate to those present in the total population


Treatment Condition: Refers to the presence of the independent variable, as opposed to the control condition in which the independent variable is absent


Treatment Group:  Subjects exposed to the levels of the independent variable; also called the experimental group


Variability: Amount of spread or dispersion within a distribution of scores