Common Research Terminology

correlation-- a type of research design that depicts a relationship between variables, but not necessarily one of cause-effect

data--information. can be numbers or words. plural form of datum. the "data show" not "shows"

dependent variable--the quality you are observing.

experiment--a research design used to find "cause-effect" relationships. the "effect of...on..". lots of variations. top shelf in research

experimental group--the one that get the treatment

external validity--how generalizable the results are outside of the study as it concerns other populations and locations.

further study is needed--we don't know

independent variable--the one you are manipulating. The effect of (independent, such as teaching method) on....

internal validity--the extent to which a study measures what it is supposed to (accuracy within the study)

mean--the arithmetic average

median--the middle where half the scores fall above, half below. eliminates the influence of outliers

mode--the score that occurs most

no evidence--we don't know, haven't figured out how to attack the problem, or haven't cared enough to try

operational definition--how a term is used in a study.

prove--not used in research about human performance. could result in a shunning

random--by chance

random sample--everybody had the same chance of being assigned to any group. sometimes confused with who you ran into by chance who would participate

reliability --the extent to which a test is repeatable with similar results

research--a systematic, objective way to find out things

research design--the game plan or method for finding out what you want to know. experiments, correlations, descriptive studies

significance--two meanings: significance of the study means why it is importance. statistical significance has a mathematical meaning

standard deviation--a measure of spread. the average deviation of a group of scores from the mean

statistical significance--an important finding that did not likely happen by chance. p<.05 means that there were less than 5 chances in 100 that the result would have happened randomly

statistics--mathematical tools based on the normal curve used to analyze data. researchers must match statistics with research designs

T-score--a standard score on the normal curve where the mean is assigned "50" with deviations of "10". Allows more simple interpretation of student achievement.

t-test--a parametric statistical tool that compares differences between the means of two groups; assumptions for use include normal distribution and at least interval data.

the extent to which--a favorite phrase of researchers that means "how much". implies ranges and probabilities and avoids absolutes

validity --accuracy. the extent to which a test or study measures what it is supposed to measure

variable --a quality of interest that can be manipulated or observed. Also see independent variable, dependent variable.

z-score--a standard score on the normal curve. Each z-score is counted as "1" from the mean, plus or minus.

Pages related to research terminology, see

Scientific Method Research

Research Designs

Top of Research Terminology

Back To Research Basics

Back to Home Page