GLOSSARY

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

 

(AKA = Also Known As)



A PRIORI Established in advance; as a method of deduction, a priori means syllogistic reasoning from propositions appearing to be self-evident



ABSCISSA The base line or X-line on a graph (the horizontal line)



ACCEPTING SAMPLE The number of responses received from a sample of subjects contacted in a survey



ACCESSIBLE POPULATION That portion of the population to which one can have access; if used, limit generalizations to this group



ACCIDENTAL SAMPLE See Sampling Error



ACTIVE VARIABLE Variables that can be manipulated and are manipulated in a particular experimental study



ALPHA LEVEL () What are the chances the results were obtained by chance? See a priori in a study. Same as Type I error



ALTERNATIVE HYPOTHESIS A hypothesis used as an alternative to the null hypothesis; also, a rival explanation or hypothesis in an ex post facto study



ANTECEDENT See Independent Variable



ASSIGNED VARIABLE See Attribute Variable



ASSUMPTIONS Propositions for which no information can be made available within the scope of the study; aka Delimitations



ASYMPTOTIC The curve lines never touch the abscissa (base)

line



ATTITUDE A predisposition to behave in a certain manner; often measured with Likert, Guttman, Thurstone or semantic differential scales



ATTRIBUTE VARIABLE Variables that cannot be manipulated; aka Assigned, Measured, Organismic, Personological, Sociological, Psycho-Sociological, and Demographic Variables



AUTHORITY Knowledge gained from one who has experience or some other expertise; sources are custom and tradition



AXIOMATIC Assume to be true without immediate proof

 

(Go to Top)



BELL-SHAPED CURVE See Normal Curve



BLIND EXPERIMENT The subjects don't know what they are getting as a treatment



BLOCKING Taking continuous data and making it into discrete data by categorizing, i.e. high/low; or matching subjects and then randomly assigning them to level of treatment



CASE STUDY Type of descriptive research with smaller sample but more depth than survey; see Naturalistic Inquiry, Ethnography, Phenomenology, Qualitative research



CATEGORICAL VARIABLE See Discrete Variable; aka Discrete or Discontinuous



CAUSAL-COMPARATIVE See ex post facto

CAUSALITY Bringing about a change in the dependent variable by manipulating the independent variables; if/then relationship



CAUSE/EFFECT Purpose of experimental research to establish causality and an if/then relationship



CHARACTERISTIC Same as variable, but this term is used in a survey or

correlational study instead of the term "variable"



CHUNK SAMPLE See Sampling Error



CLINICAL MODEL An exception to direct route/indirect route inference; it doesn't

study all people, but it takes the ones that are there at the time;

results can be generalized (have external validity) because (when) the study is replicated



CODING Marking data collection devices in a way that nonrespondents can be identified; also, placing numeric values on information gathered



COEFFICIENT OF STABILITY Consistency of subjects' scores on a measure over time; aka test-retest reliability coefficient



COEFFICIENT OF EQUIVALENCY Reliability coefficient for two forms of an instrument administered at essentially the same time (in immediate succession)



COEFFICIENT OF INTERNAL In a split-half reliability test the two

CONSISTENCY equivalent forms and contained within a single test and reflect fluctuations from one item sample to another (split half, odd-even, Kuder-Richardson and Cronbach's alpha)



COHORT STUDY A group of people belonging together is followed over a period of years and sampled at each data collection point; a random sample is drawn each time from the frame; see also Developmental Study



COMPARATIVE HYPOTHESIS Typical of a survey to see if two groups are significantly different on a domain



COMPARISON GROUP Group receiving an alternate form of the treatment; get something



CONCEPT An abstraction from observed events; e.g., liquid



CONCLUSION A proposition inferred from a finding or premise in research; is directly supported by the data



CONCURRENT CORRELATIONAL STUDY The relationship is drawn on two characteristics of the same group from data gathered at the same point in time



CONCURRENT VALIDITY Correlation between test scores and a criterion measure at the same or a very close point in time



CONFOUNDING VARIABLE See Extraneous Variable



CONSEQUENCE See Dependent Variable



CONSTITUTIVE DEFINITION The general meaning a concept or construct is supposed to have; the dictionary meaning



CONSTANT Something which does not change



CONSTRUCT Higher-level abstraction from things that cannot be observed or illustrated by specific objects or events; e.g., justice, leadership, reading readiness



CONSTRUCT VALIDITY The extent to which a test measures a specific psychological trait



CONTAMINATING VARIABLE See Extraneous Variable



CONTEMPORARY HISTORY See History Threat



CONTENT VALIDITY The extent to which the instrument represents the content of interest; i.e., established by a panel of experts



CONTINUOUS VARIABLE A variable that can take on any value along a scale, spectrum or continuum



CONTROL Purpose or end sought of experimental and quasi-experimental research to test hypotheses of causal relationships between variables achieved by the design of the study



CONTROL GROUP The subjects at one level of treatment in the study who get nothing



CORRELATIONAL STUDY Type of relational study to explain and predict; compares two or more different characteristics from the same group of people



COUNTERBALANCE DESIGN Quasi-experimental design where all Ss get treatments to find out the best sequence, aka Rotational Experiment, Cross-Over Design, Switch-Over Design, and Latin Square



CRITERION See Dependent Variable



CRITERION-RELATED VALIDITY The relationship between the scores on a measuring instrument and a criterion believed to measure directly the behavior or characteristic in question



CRONBACH'S ALPHA A formula for calculating internal consistency reliability



CROSS-OVER DESIGN See Counterbalance Design

 

(Go to Top)



DASHED LINE An experimental group and comparison group are equated by a system other than random assignment; e.g., matching/blocking



DATA SAMPLE The number of usable responses received from subjects in a survey



DEDUCTION Syllogistic reasoning; determines the consequences of pre-existing knowledge; theorizes about facts



DELIMITATIONS Basic assumptions of a study; see Assumptions



DEMOGRAPHIC VARIABLE See Assigned Variable



DEPENDENT VARIABLE The phenomenon that is the object of the study which is

going to be affected, aka Consequence, Criterion,

Observation, Y, the thing that is supposed to change

because of X



DESCRIBE Purpose of descriptive research used to portray the incidence, distribution, and characteristics of a group or situation



DESCRIPTIVE RESEARCH Type of research used to explore and describe; asks "what exists?"



DEVELOPMENTAL STUDY Type of descriptive research taken over a period of time; begins now and continues into the future; See Follow-Up Study; aka Longitudinal Study, Trend Study, Cohort Study, Panel Study



DICHOTOMOUS VARIABLE Two categories of a discrete variable; mutually

exclusive categories



DIFFERENTIAL MORTALITY Error occurring when subjects drop out of the groups in different levels of a treatment at different rates



DIFFERENTIAL SELECTION See Selection Bias Threat, selecting members of experimental groups differently



DIRECT ROUTE INFERENCE See Perfect Induction



DIRECTIONAL TEST See One-Tailed Test



DISCONTINUOUS VARIABLE See Discrete Variable; aka Discrete or Categorical



DISCRETE VARIABLE Can only take on a specific value, aka Categorical and Discontinuous Variables



DOMAIN The area or thing of interest, e.g., numerous items on a questionnaire might measure the domain of attitude toward school



DOUBLE-DIPPING Used to control non-response error in surveys; select a random sample of non-respondents, get their responses, and statistically compare to the respondents

(Go to Top)

ECOLOGICAL VALIDITY Conclusions will hold true for all similar situations; external validity



EFFICACY Efficiency or goodness of a study



EMPIRICAL RESEARCH Researcher develops hypotheses based on the findings of related research and conducts study in a systematic manner, aka scientific



EMPIRICISM Method of scientific inquiry



END SOUGHT The aim or purpose of a research study



ETHNOGRAPHY The scientific description and classification of the various cultural and racial groups of humankind



EX POST FACTO STUDY A type of relational study that does not establish causality but is often called causal-comparative and substitutes for an experimental study; its purpose is to explain and predict; subjects self-select the level of X or variable (X) is naturally occurring



EXPERIMENTAL GROUP The subjects of a study that get the treatment (X) of interest versus the control group who do not



EXPERIMENTAL MORTALITY See Mortality Threat



EXPERIMENTAL STUDY Type of research to establish causality, if/then; cause/effect relationships; its purpose is to control



EXPLAIN In correlational research, show how two characteristics vary together.



EXPLORE Purpose of descriptive research to become familiar with existing phenomena



EXTERNAL VALIDITY To whom we can generalize the results; i.e., populations and environments (Ecological plus Population Validity)



EXTRANEOUS VARIABLE One variable you should have controlled, but didn't, and one which contaminates the results of the study; aka Contaminating, Confounding, Intervening



EXTRAPOLATION Make a generalization about the results beyond those subjects studied

(Go to Top)

FACE VALIDITY Content validity determined by the researcher; or "looks" valid to the subjects



FACTOR See Independent Variable



FIELD TEST Used to determine suitability/ utility/clarity of an instrument



FINDING Result of the study relative to the problem



FOLLOW-UP STUDY A type of developmental study which goes back in time a number of years versus a longitudinal study which goes forward in time



FRAME List of members in a population



FRAME ERROR Discrepancy between the intended target population and the actual population from which the sample is drawn



FUNCTIONAL RELATIONSHIP Relationship in which it has been determined that a change in one variable (X) is accompanied by a change in the other (Y), based on complex interaction rather than direct cause



FUNCTIONAL THEORY Type of theory based on applied research and direct observation, aka Nominal Theory, Normative Theory

 

(Go to Top)

Ha A schematic representation of an alternative hypothesis; see Alternative Hypothesis; same as H1



Ho A schematic representation of the null hypothesis; see Null Hypothesis



H1 A schematic representation of an alternative hypothesis; see Alternative hypothesis; same as Ha



H2 A schematic representation of a second alternative hypothesis (same as Ha')



HETEROSCEDASTIC There is no uniformity of the data distribution on a scatterplot; e.g., Y has greater variability at some points on X than others



HISTORICAL RESEARCH Draws conclusions about the past



HISTORY THREAT Things occurring between the first and second measurement (or during X) besides the treatment variable, aka Contemporary History or Unique Intrasession History



HOMOSCEDASTIC The spread of the data distance is about the same on a scatterplot; i.e., variability of Y is uniform across values of X



HUMAN SUBJECTS REVIEW Subjects of a study are protected; they must know all limits of confidentiality



HYPOTHESIS A tentative solution to a problem or answer to a question; it becomes or contributes to a theory once it is tested; plural = hypotheses



HYPOTHETICAL-DEDUCTIVE THEORY Series of deduced consequences that are logically derived from the hypothetical statements; if...,then...

 

(Go to Top)

IMPERFECT INDUCTION Examining examples (sample) and inferring to the rest (population) with probability; i.e., using a sample to generalize to a population like indirect route inferences



IMPLICATION The relation which holds between two propositions because one is logically deducible from the other



INDEPENDENT VARIABLE Variable predicted from and antecedent to the dependent variable; aka Antecedent, X, Treatment, Factor



INDIRECT ROUTE INFERENCE The typical way to survey groups because we cannot observe everyone in the population; uses imperfect induction; sampling



INDUCTION General conclusions on the basis of knowledge gathered through direct observation, liked to Francis Bacon; observes and accumulates facts



INDUCTIVE THEORY Summarizing theory that emerges after careful consideration of the facts



INFERENTIAL THEORY See Inductive Theory



INFERENTIAL STATISTICS Tools to see if results are due to chance; see also Alpha Level



INSTRUMENTATION THREAT A change in calibration, observers, tests, or scorers used may produce a change in measurements and bias results



INSTRUMENT A device for measuring the value of the variables (characteristics) under observation



INTERACTION OF SELECTION & X Subjects self-select, and they want to help the experimenter; external validity threat--can only generalize to volunteers



INTERACTION OF TESTING How can you generalize to people who WITH TREATMENT weren't pretested if all in the experiment were pretested; external validity threat



INTERNAL CONSISTENCY Reliability which establishes that each item contributes to the measurement of the domain of interest



INTERNAL VALIDITY In an experiment, did the treatments make a difference in the specific experimental instance or was it something else?



INTER-RATER RELIABILITY Consistently reliable across several observers



INTERVENING VARIABLE See Extraneous Variable



INTRA-RATER RELIABILITY Consistent reliability on different occasions with the same observer


(Go to Top)


KUDER-RICHARDSON Formulas for calculating internal consistency reliability often called KR-20 and KR-21

 


LATIN SQUARE DESIGN See Counterbalance Design



LAW Statement of a relation or sequence of phenomena which does not vary under the same conditions



LEVEL Values of the independent variable or treatment administered



LIMITATIONS Procedural Factors which impose limits on the

study; included in Chapter I of the research

proposal



LONGITUDINAL STUDY See Developmental Study

 

(Go to Top)

MAIN HYPOTHESIS The presumed relationship between variables to be measured in the study; principal X's in ex post facto research; hypothesis written for a main

independent variable



MANIPULATION Random assignment of experimental units (subjects) to levels of the independent variable (treatment)



MATURATION THREAT Processes within respondents operating during X as a function of the passage of time which would affect Y, e.g. getting older/more mature





MAXMINCON PRINCIPLE Maximize the variance of the levels of the variables in the hypothesis, minimize error or random variance including errors in measure- ment (reliability and validity), and control unwanted contaminating variables that have an effect on experimental outcomes (See Kerlinger)



MEASURED VARIABLE See Attribute Variable



MEASUREMENT Assigning numbers to phenomena according to a set of rules so that they reflect the real world



MEASUREMENT ERROR The data gathered must be true, reliable, and valid; changes in measuring instruments, scorers, or observers may produce changes in results which aren't due to X



METHODOLOGY See Procedures



MODEL The analogy, conceptual scheme, or structure that could explain the relationships of variables under consideration; a type of theory; aka Paradigm



MORTALITY THREAT Loss of subjects during the study; particularly a problem if more subjects lost at one level of X than the other(s) (Differential Mortality)



MULTIPLE TIME SERIES DESIGN Quasi-experimental design; attempt to match with a similar group who does not receive the treatment; like time series and nonequivalent control group designs



MULTIPLE-TREATMENT Previous treatments cannot be erased; INTERFERENCE external validity threat

 

(Go to Top)

NATURALISTIC INQUIRY The assumption by scientists that the events in nature are orderly and regularity can be discovered through the scientific method, aka Case Study, Ethnography, Phenomenology, Qualitative Research, Grounded Theory



NEGATIVE CORRELATION When one variable increases, the other goes down on a scatterplot, producing a negative correlation



NOMINAL THEORY See Functional Theory



NONDELIVERABLES Data collection devices mailed but returned because of wrong names and/or addresses in the frame



NONDIRECTIONAL TEST See Two-Tailed Test



NONEQUIVALENT CONTROL Quasi-experimental design often used in GROUP DESIGN educational research; refinement of one group pretest-posttest design; has no randomization



NONRESPONDENTS Individuals from whom data have not been collected



NONRESPONSE ERROR Subject cannot be located or fails to respond, refuses, or does not return the mailed questionnaire, phone call or interview



NONSIGNIFICANT Results could be due to chance; accept Ho and reject alternative hypothesis(es)



NORMAL CURVE Symmetrical distribution of measures with the same number of cases at specified distances below the mean and above the mean, aka Bell-Shaped Curve



NORMATIVE STUDY See Survey Research; aka Status Study



NORMATIVE THEORY See Functional Theory



NULL HYPOTHESES Hypotheses of "no difference" or "no relationship", aka Statistical Hypotheses; not always stated in proposal

 

(Go to Top)

OBSERVATION See Dependent Variable; comprised of true score plus error; i.e., O = TS + E



ONE-TAILED TEST Direction of the result is predicted in the hypothesis, either < or >



ODD-EVEN A type of internal consistency reliability

OPERATIONAL DEFINITION The meaning of a word as it will be used in the study; how the concept or construct will be measured or manipulated



ORDINATE The vertical line or Y-line of a graph (or scatterplot)



ORGANISMIC VARIABLE See Attribute Variable

 

(Go to Top)

PANEL STUDY A sample is selected from the population at an initial data collection point, and the same sample is used at each data collection point; allows us to look at changes in individuals as well as the group; see also Developmental Study



PARADIGM See Model



PARAMETERS Characteristics of a population, represented by Greek letters; there are no sampling errors and no statistics used



PARSIMONIOUS Economical or frugal, condensing large number of things into smaller groups



PATH ANALYSIS Statistical tool used to analyze relations within a model to see how some characteristics relate to others



PEARSON PRODUCT MOMENT Type of correlational statistical tool

CORRELATION assuming linearity and homoscedasticity and both X and Y to be continuous (interval) data



PERFECT INDUCTION Observing all examples (members) of a population; Direct Route Inference; Census



PERFECT PREDICTION Condition resulting when a correlation = +/- 1.00



PERSONOLOGICAL VARIABLE See Attribute Variable



PHENOMENOLOGY The science of facts, occurrences, or circumstances observed or observable, aka Case Study, Naturalistic Inquiry, Ethnography



PHILOSOPHICAL RESEARCH Asks value questions



PILOT TEST Used to determine reliability with data gathered from Ss like those to be studied



POPULATION All members of any well-defined class of people, events, or objects



POPULATION VALIDITY To whom the results can be generalized; external validity



POSIT To state a hypothesis



POSITIVE CORRELATION When one characteristic (variable) goes up (increases), so does the other



POWER A good statistical decision; it does not mean goodness; directly related to Type I and II errors



PRACTICAL SIGNIFICANCE With direct observation it doesn't look like there is a meaningful difference even if statistically significant, some statistics help here like r2, Omega2 and Eta2



PREDICT One variable (Y) forecasted from knowledge of the other (X); see also Regression



PREDICTIVE CORRELATIONAL Knowing an earlier set of data on a STUDIES characteristic can lead to forecasting a

later set or other characteristic



PREDICTIVE VALIDITY Correlation between test scores and a criterion that occurs at a later point in time



PRE-EXPERIMENTAL DESIGN Characterized by minimal control and many threats to validity; see the Campbell and Stanley book



PREMISE A proposition from which a conclusion is drawn



PRETEST SENSITIZATION See Testing Threat



PROBABILITY The odds or chances of an event occurring



PROBLEM SOLVING Reflective thinking using the scientific approach



PROCEDURES Chapter of proposal, aka Methods, Chapter III



PROGRAMMATIC RESEARCH Finding an area of expertise in one's

EMPHASIS discipline in which you conduct numerous research studies



PSYCHO-SOCIOLOGICAL See Attribute Variable

VARIABLE



PURPOSIVE SAMPLE A nonrandom (nonprobabilistic) sample, a group study which was "handy"; not scientific, aka:

Accidental sample, "Chunk" Sample

 

(Go to Top)

QUALITATIVE RESEARCH See Case Study



QUASI-EXPERIMENTAL Controls are stronger than pre-

DESIGN experimental but not as strong as true experimental; used because of necessity, not lack of knowledge



QUESTIONNAIRE Method of data collection for descriptive re- search using a list of questions; often mailed


(Go to Top)


R F P Request for proposal



RANDOM ASSIGNMENT See Randomization; Manipulation



RANDOM SELECTION The use of a chance (probabilistic) procedure to select a sample from a population



RANDOMIZATION The assignment of subjects to groups in such a way that for any given placement, every member of the population has an equal probability of being assigned to any of the groups; the most powerful tool in research, Noted as "R";

Same as Manipulation; Active Independent Variable



RATIONALISM Method of intuition



REACTIVE EFFECTS OF Person is exposed to the treatment in a non

EXPERIMENTAL experimental environment; interaction of history ARRANGEMENTS and treatment; external validity threat; one treatment affecting a subsequent treatment



REACTIVE TEST Administration of one test helps performance on another test



REGRESSION Predictability; if you know one variable, how well you can predict another; a set of statistical tools



RELATIONAL RESEARCH Type of research which asks if two characteristics vary together (explain) and asks how well Y can be predicted from X (predict); see Correlational Study and Ex Post Facto Study



RELIABILITY Does the instrument or technique measure consistently whatever it measures?



REPLICATION Repeating, redoing, or reproducing a study



RESEARCH A formal, systematic application of the scientific approach to the study of a problem to discover new information or expand and verify existing knowledge



RESEARCH HYPOTHESES Hypotheses derived from theory and/or related research, logical argument, or the opinions of others; stated in the proposal; the Ha (alternative hypothesis) in prose form; aka Scientific Hypotheses and Substantive Hypotheses



RIVAL HYPOTHESIS Possible alternative explanation for the phenomenon which should be tested in ex post facto research to assure that X did make the difference because you cannot randomly assign subjects



ROTATIONAL EXPERIMENT See Counterbalance Design

 

(Go to Top)

SAMPLE The randomly selected group that is observed in a study



SAMPLING ERROR Obtaining a nonprobabilistic group of subjects from a population; e.g., Accidental Sample, Chunk Sample, or Purposive Sample



SCALES Devices for measurement placed on a continuum or in a category



SCIENTIFIC HYPOTHESES See Research Hypotheses



SCIENTIFIC PROCESS Chain of reasoning in which the investigator moves inductively from observation to hypo- thesis, then deductively from the hypothesis to the logical implication of the hypothesis



SELECTION BIAS THREAT Varying the selection of comparison groups, aka Differential Selection



SELECTION ERROR Type of internal validity threat where some members of a group had a greater chance of being selected than others (on a list, frame, twice)



SELECTION-MATURATION THREAT Seen in quasi-experimental designs where intact groups have different rates of maturation and were selected differently



SHOTGUN EMPIRICISM Going out and gathering large quantities of data and then using high-powered statistical tools to get something meaningful from all the data; not good scholarship, or studying variables without a theoretical bases



SLOPE The angle of the regression line tells how well you can predict



SOCIOLOGICAL VARIABLE See Attribute Variable



SPLIT-HALF A type of internal consistency reliability



SPURIOUS CORRELATION Occurs by chance although it may be a high correlation



STATISTIC Characteristics of a sample, symbols written in Latin or English letters



STATISTICAL HYPOTHESES See Null Hypotheses



STATISTICAL SIGNIFICANCE Differences are not due to chance, see alpha level; reject Ho and accept H1 (Ha) or H2 (Ha')



STATISTICAL REGRESSION THREAT Groups selected on the basis of extreme scores on an initial measure will move toward the mean of the group of the next measurement; the most subtle threat to internal validity



STATUS STUDY See Survey Research; aka Normative Study; Survey Study



STRATIFICATION Selecting a sample with the population divided into some subgroups (strata)



SUBSTANTIVE HYPOTHESES See Research Hypotheses



SUITABILITY Can people read, understand and respond to the instrument or technique



SURVEY RESEARCH Type of descriptive research, aka Status Studies and Normative Studies



SWITCH-OVER DESIGN See Counterbalance Design



SYLLOGISM Major premise, minor premise, and a conclusion; see also Deduction

 

(Go to Top)

TARGET POPULATION Identification of the population to be represented in the study; group to whom results can be generalized; see also Frame; aka Universe



TENACITY Tradition; always has been true, a way of obtaining knowledge



TEST-RETEST A test given twice to the same individuals over time to determine reliability, coefficient of stability is produced



TESTING THREAT Effects of taking a test upon the score on the second test, aka Pretest Sensitization



THEORY Summarizes and puts into order the existing knowledge in a particular area



TIME SERIES DESIGN Quasi-experimental design using several pre-observations, the treatment, and several post-observations



TRADITION See Tenacity



TREATMENT See Independent Variable



TREND STUDY General populations are sampled at each collection point; same individuals aren't sampled; every year the frame changes; see also Developmental Study



TRIANGULATION Using more than one technique of data collection (e.g., survey and case study) thus making the results more convincing and thorough



TWO-TAILED TEST No direction is given in the hypothesis; it can be > and < or



TYPE I ERROR Probability of rejecting the hypothesis even though it is true; see alpha level



TYPE II ERROR Probability of retaining the hypothesis even though it is false

 

(Go to Top)

UNIQUE INTRASESSION See History Threat



UNIVERSAL DETERMINISM The scientific belief that all natural phenomena have antecedent factors



UNIVERSE See Target Population


(Go to Top)


VALIDITY Does the instrument or technique measure what it purports to measure



VARIABLE Characteristic by which a person or object can be described



VARIATE Specific value on a variable

 

(Go to Top)

WORKHORSE DESIGN Pretest-posttest control group experimental design



X See Independent Variable



Y See Dependent Variable

 

(Go to Top)