Chapter 5 slides

Chapter 5 slides - Chapter 5 Chapter Measurement Concepts...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Chapter 5 Chapter Measurement Concepts Theory of Measurement Theory Reliability The level of consistency or stability in our measures The of behavior of X=T+E Observed Score (X) True Score (T)- a theoretical real score on a variable True that exists with no error. that Measurement Error (E)- the inaccuracy that exists Measurement when measuring a psychological characteristic or behavior. It gives us an indication of reliability because we know that less error is a sign of a more reliable test. reliable Types of Reliability Types Test-Retest reliability Measuring the same individuals at 2 points in time. Carry-over Measuring effects are possible. effects Internal Consistency reliability Split-half reliability Comparing 1st half of test with 2nd half Odd-even split Comparing odd items with even items Cronbach’s α Cronbach’s Average of correlation between items Interrater reliability Extent to which two or more “raters” agree Used often in studies of infant behavior Accuracy of Measures Accuracy Are you really measuring what you think Are you’re measuring? you’re Construct validity Accuracy = Validity (more or less…) Reliability vs. Validity Reliability Reliability is necessary but not sufficient Reliability for validity for If reliable If unreliable If valid If not valid Reliability vs. Validity Reliability Reliability is necessary but not sufficient Reliability for validity for If reliable may or may not be valid If unreliable If valid If not valid Reliability vs. Validity Reliability Reliability is necessary but not sufficient Reliability for validity for If reliable may or may not be valid If unreliable NOT valid If valid If not valid Reliability vs. Validity Reliability Reliability is necessary but not sufficient Reliability for validity for If reliable may or may not be valid If unreliable NOT valid If valid reliable If not valid Reliability vs. Validity Reliability Reliability is necessary but not sufficient Reliability for validity for If reliable may or may not be valid If unreliable NOT valid If valid reliable If not valid may or may not be reliable Indicators of Construct Validity Validity Face validity Does the content of the measure actually Does measure the variable? measure Given the definition of the variable, does it look Given like you’re measuring it. like Predictive validity Does your measure allow you to predict the Does behaviors that it should predict? SAT/ACT tests Concurrent validity Do two or more groups differ in an expected Do way? way? Convergent validity Do your findings match other researchers’ Do findings measuring the same construct? findings Discriminant validity Measures are not related to variables with Measures which it should not be related which Being unobtrusive Being The experimenter can actually bias the The reliability and validity of his or her measure. measure. Reactivity Awareness of being measured changes an Awareness individual’s behavior. individual’s How to combat this: Get creative and find ways to indirectly study the Get variable of interest. variable Categorical vs. Continuous Variables* Variables* Categorical variables Has distinct boundaries between items Continuous variables No distinct boundaries; a gradient of values No exists for the variable exists Not in the textbook in this format. Scales of Measurement Scales Nominal Scale Categorical groups (a.k.a. the Naming Scale). Coding nominal data merely indicates a difference in Coding category, class, quality, or kind. category, Nominal variables do not provide meaningfully Nominal ordered, numerical scores ordered, Items used with this scale can be put into any order. Example: Catholic, Protestant, Jew, Muslim, Buddhist, Other Ordinal Scale Ranking items Has named categories or numerical scores with Has additional property of allowing categories to be ranked from highest to lowest (or any other order you choose). you This does not show the distance between items in This the scale. the Example A bicycle race Interval Scale Numbers on the scale are of equal size Has all the characteristics of nominal and Has ordinal scales. ordinal There is no absolute zero Example Fahrenheit and Celsius thermometers Ratio Scale Has intervals of equal size and a true zero Has point point We can compute ratios (the amount of one We observation in relation to another) observation Example Height and Weight Activity Question Activity Identify which scale should be used: The temperatures in cities throughout the The country that are listed in most newspapers. country The birth weights of babies who were born The last week. last The number of hours you spent studying for The your last test. your The amount of tip left after each meal at a The restaurant during a 3-hour period. restaurant The number of votes received by the The Republican candidates for Governor in the recent election. the The brand listed third in a consumer The magazine’s ranking of DVD players. magazine’s Your friend’s score on an intelligence Your test. test. The wall color in my office is white and The my boss’s office is brown. my The type of programming on each radio The station in your city. (e.g., KPSY plays jazz, KSOC is talk radio) jazz, ...
View Full Document

This note was uploaded on 06/06/2011 for the course PSYC 226 taught by Professor Hutcheson during the Spring '11 term at South Carolina.

Ask a homework question - tutors are online