学前教育研究方法chapter--10--Measurement--Research.ppt

CHAPTER 10Measurement ResearchvThe areas of educational measurement,indeed the measurement of educational variables such as achievement and attitude,are of themselves disciplines of study.It is the intent of this discussion to provide an overview of measurement as it is a part of conducting research,including basic measurement concepts.There are suggestions about sources of tests and other measurement instruments,sources that and be very helpful in locating examples of the more commonly used measurement instruments in educational research.1v1.Concepts of measurementvMeasurement is a process of assigning numerals according to rules.The numerals are assigned to events or objects,such as responses to items or to certain observed behaviors.2v2.Types of Measurement ScalesvThe four general levels of measurement scalesnominal,ordinal,interval,and ratio-were defined in Chapter 2 and briefly reviewed here.The four scales comprise a hierarchy of measurement levels based on the amount of information contained in the score or the measure generated by the scales.The scales to from nominal to ratio in order from least to the most information contained.The four scales are defined as follows;in each case,an example of a variable that could be measured by the type of scale is given:3vNominal:This gives categorization without order,whatever is being measured is categorized into two or more classifications that indicate only differences with respect to one or more characteristics.Example:gender of the individual.4vOrdinal:In addition to indication difference,this scale also orders the scores on equal intervals between scores are not established.Example:attitude toward school.5vInterval(also called equal unit):In addition to order,equal units or intervals are established in the scale such that a difference of a point in one part of the scale is equivalent to a difference of one point in any other part of the scale.Example:temperature.6vRatio:In addition to an equal unit,the scale contains a true zero point that indicates a total absence of whatever is being measured.Example:monetary expenditures for various school functions.7v3.Reliability of MeasurementvTwo essential characteristics of measurement that must be considered in establishing the appropriateness and usefulness of measurement instruments are reliability and validity.vReliability means consistency of the instrument in measuring whatever it measures.8vReliability coefficients can take on take on values from 0 to 1.0,inclusive.Conceptually,if a reliability coefficient were 0,there would be no“true”component in the observed score.The observed score would consist entirely of error.On the other hand,if the reliability coefficient were 1.0,the observed score would contain no error;it would consist entirely of the true score.Clearly,in educational measurement,it is desirable to obtain high reliability coefficients,although coefficients of 1.0 are very rare indeed.9vEquivalence ReliabilityvParallel forms or alternate forms:This procedure involves the use of two or more equivalent forms of the test.The two forms are administered to a group of individuals with a short time interval between the administrations.If the test is reliable,the patterns of scores for individuals should be about the same for the two forms of the test.There would be a high positive association between the scores.10vStability ReliabilityvTest-retest:In this procedure,the same test is administered on two of more occasions to the same individuals.Again,if the test is reliable,there will be a high positive association between the scores.11vInternal Consistency ReliabilityvSplit-half:This procedure requires only one administration of the test.In computing split-half reliability,the test items are divided into two halves,with the items of the two halves matched on content and difficulty,and the halves are then scored independently.If the test is reliable,the scores on the two halves have a high positive association.An individual scoring high on one half would tend to score high on the other half,and vice versa.12vKuder-Richardson procedures:Two formulas for estimating reliability,developed by kuder and Richardson(1937),require only one administration of a test.One formula,KR-20,provides the mean of all possible split-half coefficients.The second formula,KR-21,may be substituted for KR-20,if it can be assumed that item difficulty levels are similar.13vCronbach alpha:A formula developed by Cronbach(1951),based on two of more parts of the test,requires only one administration of the test.14v4.Validity of MeasurementvThere are basically two approaches to determining the validity of an instrument.One is through a logical analysis of content of a logical analysis of what would make up an educational trait,construct,or characteristic.This is essentially a judgmental analysis.15vThe other approach,through an empirical analysis,uses criterion measurement,the criterion being some sort of standard or desired outcome.The criterion measure might be performance on a task or test,or it could be a measure such as job performance.Validity is then a measure of the association or correlation between the test being validated and the criterion measure.16vContent-Related EvidencevContent validation is the process of establishing the representativeness of the items with respect to the domain of skills,tasks,knowledge,and so forth of whatever is being measured.Thus,content-related evidence deals with the adequacy of content sampling.Content validation is a logical analysis of the items,determining their representativeness.Validity of achievement tests is commonly based on content-related evidence.17vCriterion-Related Evidence:Concurrent and Predictive.vCriterion validation establishes validity through a comparison with some criterion external to the test.The criterion is,in essence,the standard by which the validity of the test will be judged.If the scores of the measure being validated relate highly to the criterion,the measure is valid.If not,the measure is not valid for the purpose for which the criterion measure is used.18vConstruct-Related Evidence.vConstruct validation can involve both logical and empirical analyses.The term construct refers to the theoretical construct or trait being measured,not to the technical construction of the test items.A construct is a postulated attribute or structure that explains some phenomenon,such as an individuals behavior.Because constructs are abstract and are not considered to be real objects or events,they sometimes are called hypothetical constructs.Theories of learning,for example,involve constructs such as motivation,intelligence,and anxiety.19vReliability is a necessary but not sufficient condition for validity.That is,a test or measuring instrument could be reliable but not valid.In that case,it would be consistently measuring something for which it was not intended.However,a test must be reliable to be valid.If it is not consistent in what it measures,it cannot be measuring that for which it was intended.If it doesnt measure something consistently,its not going to be useful or valid.20