Stats Vocabulary.docx

Stanford Vocabulary, Fall QuarterEDUC200ADescriptive The nature of subjects’ characteristics determines which particular subjects end up in each comparison groupExperimental A type of research used to establish cause-and-effect relationships by manipulating what subjects will experience.Non-experimental: Correlational A type of research that measures two or more variables and looks to see how the variables are related to each other without any direct manipulation or direct control of conditions experienced.Causal To establish causationResearch A systematic and scientific approach to finding answers to questionsResearch Design A plan for gathering data for answering specific research questionsStatistics The methods used on the data collected to answer the research questionsDescriptive Statistics Methods used to obtain indices that characterize or summarize data collected from a particular sampleInferential Statistics Methods that allow the researcher to make inferences from a set of data collected from a sample to a larger populationResearch Hypothesis A tentative statement about the expected relationship between two or more variablesInternal Validity Validity (true for every circumstance) of causal inferences (conclusions from premises known/assumed to be true)External Validity The extent to which the results of a study can be generalized to other situations/peopleDifferences between internal and external validity: Internal (was it really the treatment that caused the difference between subjects) and external (can findings be generalized?)Hawthorne Effect Changes in participant behavior during a study due to social situations and treatment they receivedDependent (Criterion) Variable The output/effectIndependent (predictor) Variable The input/causesFallacy of homogeneity NEED TO DEFINEPost Hoc fallacy A occurs before B, therefore A is the cause of BBlind Observers Person gathering the data is unaware of the treatment being applied to each subjectStatistical significance Statistical assessment of whether observations reflect patterns, not chanceUnivariate/Multivariate One/More than one variableParametric Statistics Assumes data has come from a type of probability distributions (assigns probability to each of the possible outcomes of a random experiment) and makes inferences about the parameters of the distributionShotgun Approach Aim of covering as wide a population/area as possibleCoefficient of Determinant A measure of the proportion of variability in a data set that is accounted for by a statistical modelVariable What is measured or varied. Must be measurable and must varyMeasurement The application of rules in assigning numbers to cases so as to represent the presence or absence of quantity of an attribute possessed by each case.Nominal Scale Something that’s a category. You must fit into oneOrdinal Scale Used to rank someone (ranked 1-x, distances between don’t have to be equal)Interval Scale The number assigned is the amount of attribute (0 is not the absence of attribute, it’s arbitrary. Distance between intervals is equal)Ratio Scale The number assigned is the amount of attribute (0 is absence of attribute)Likert-Scale Data Rating a construct/behavior from 1-5Pre-Experimental Design Doesn’t have an adequate control group AND/OR random assignmentOne-Shot Case Study A treatment is administered to a group of subjects and an observation is made to measure effect of the treatmentOne-Group Pretest-Posttest Design Group of subjects receives a pretest, treatment, and posttestIntact Group Comparison Design Two existing groups of subjects (one control, one treatment) receive treatment and are observed at the endExperimental Design Random assignment, intervention, existence of control group, and observation of resultsPosttest Only Experimental Group Subjects are randomly assigned to control and experimental group, receive identical treatments (except for IV) and take a posttestPretest-Posttest Control Group Design Subjects randomly assigned to C and E group, receive pretest, identical treatments (except IV), and are given a posttestFactorial Design Two or more IV are manipulated at the same time (each has two levels). Each level is combined with the other level variables, making four groups. Subjects are randomly assigned and measured on the DVQuasi-Experimental Design Researcher unable to randomly assign. Has two comparative groupsNonequivalent Control Group Design Those in C and E group receive a pretest, treatments (except for IV), and a posttest (not randomly assigned)Time Series Design Multiple observations are taken at equally spaced periods of time before treatment is implemented. After treatment, observations are taken again at periods of time (prior to experiment, subjects are control, after, they’re experimental group)Ex Post Facto Situation already going on, gathering information, looking for association between variable and outcomesCorrelational Design One group of subjects is measured on t。