Chapter Five Identifying Good Measurement 1 What is operationalization and why do we want to operationalize variables Give two examples of variables and their operational definitions Operationalization is the process of turning a concept of interest into a measured or manipulated variable We want to use operational definitions because they provide definitions of variables that are objective and can be standardized for everyone Example 1 Happiness could be operationally defined as subjective well being Example 2 Obesity could be operationally defined as individuals within the 95th percentile of Body Mass Index BMI values 2 What are the three most common types of measurement What are the benefits and drawbacks of each 1 Self report measure The self report measure operationalizes a variable by recording people s answers to questions about themselves in a questionnaire or interview Benefits Quick and easy Often provide the best most meaningful data Can provide info that cannot be attained by any other measures They are great measures of what people think is influencing their behavior Drawbacks The reasons for a behavior that are reported are usually inaccurate people know what they want or feel but not why People s reported memories are not as accurate as they think People s confidence in the accuracy of their memories is virtually unrelated to how accurate their memories actually are 2 Observational behavioral measure The observational measure operationalizes a variable by recording observable behaviors or physical traces of behaviors When a researcher watches people or animals and systematically records what they are doing Benefits People cannot always state the true reasons for their behavior or report on past events accurately They are great measures of what actually influences people s behavior Can be the basis for frequency claims Can be used to operationalize variables in associative and causal claims Good construct validity Drawbacks Observer bias is a potential threat to construct validity Observers see what they expect to see Observers can affect what they see biases can change the behavior of those they re observing The observer might react to being watched Acting differently than they normally would without an observer present Observation could be unethical Check with IRB 3 Physiological measure A physiological measure operationalizes a variable by recording biological data such as brain activity hormone levels or heart rate Usually require equipment to record measure and analyze biological activity Benefits Precise nominal data Can measure unique constructs Drawbacks Usually require equipment expensive Could be flawed lacking construct validity not measuring what they intend to measure Ex Thinking head circumference could be used to measure intelligence Based off idea that smarter brains are in larger skulls 3 Be able to distinguish the two types of measurement scales categorical quantitative Give two examples each Categorical measurement scales measure categorical nominal variables by separating them into categories with no numerical meaning Examples Sex levels categories would be male and female Species Quantitative measurement scales measure levels of quantitative variables that are coded represented by meaningful numbers Quantitative variables are measured in numbers Examples height weight IQ 4 What are three distinct types of quantitative variables Give 2 examples of each 1 Ordinal variables are quantitative variables with numerals representing rank order We know which are better but not how much better they are Examples Ranking a hotel by 1 2 3 4 or 5 stars Ranking 5 movies 1 5 from most enjoyable to least enjoyable 2 Interval variables are quantitative variables with numerals that represent equal intervals between levels AND have no true zero Examples Temperature in degrees Celsius 0 does not mean there is no temperature IQ score test of 0 does not mean someone has no intelligence 3 Ratio variables are quantitative variables where the numerals have equal distance and where zero means nothing Examples Brain activity zero means no increase Income 0 is no money 5 What are the three types of reliability discussed in class Give an appropriate situation for utilizing each type of reliability i e when would you use each 1 Test Retest Reliability means that he pattern of results should be the same if you give a test and then retest the same sample later on This shows the consistency of results Can be used for self report observational or physiological constructs but are most useful when measuring a construct that you expect to be stable over time Example situation Testing IQ one week and again a week later 2 Interrater Reliability is when two or more independent observers will come up with the same findings Best used for observational measures Example situation When observing how many times a person smiles while watching a 2 minute video both observers should have the same number recorded for the same person participant 3 Internal Reliability is established when the pattern of responses is consistent across items measuring the same construct Example situation When two items evaluating aggression have consistent results regardless of differences in wording or order 6 What type of graph can be used to display information about reliability Scatterplot 7 What type of coefficient can be used to evaluate reliability How do we know if there is good reliability or weak reliability What aspects of the coefficient should we pay attention to Correlation coefficient r is used to evaluate reliability and indicates how close the dots on a scatterplot are to a line drawn through them The two aspects of the coefficient are slope direction and strength Slope is direction of the relationship and can be positive negative or zero sloping up down or not at all Strength refers to how spread out the points are If the points dots are close to the line then the relationship is strong If the points dots are spread out it is weak The correlation coefficient r tells you the direction and strength of the correlation factors that are used by psychologists in evaluating reliability evidence Strong correlations are closer to 1 the strongest possible positive relationship or 1 the strongest possible negative relationship 8 What is Cronbach s Alpha How is it calculated conceptually Cronbach s Alpha is a correlation based statistic used to determine if measurement scales have internal validity It is
View Full Document
Unlocking...