PSY3213C Research Methods in Psychology Exam 2 Study Guide Chapters 5 8 Chapter 5 Identifying Good Measurement Review make sure you understand the challenges associated with operationalizing variables It s difficult to operationalize variables because often conceptual definitions can be operationalized in multiple ways with multiple components It s important to try and cover all of the components and be specific o What are the advantages disadvantages of using self report observational physiological measures to operationalize a variable ratio Self report measures Participants report on their own behavior or state of mind They are popular and easy to use however you cannot be sure that a participant is being untruthful biased or subjective Observational measures record observable behaviors or physical traces of behaviors Either frequency amount of times a behavior occurs or latency reaction time before behavior occurs May not allow you to determine the underlying cause for behavior Physiological measures Record biological data such as heart rate and blood pressure Usually noninvasive and allow you to make precise measurements of arousal of a subject s body Understand the 4 scales of measurement nominal ordinal interval Nominal categorical Events are assigned to categories Gender M F Hair color brown blonde etc Ordinal permits events or scores to be rank ordered 1st 2nd 3rd places The difference between 1st and 2nd is not necessarily the same difference between 2nd and 3rd Like in a race Interval Permits rank ordering but with equal intervals Difference between 34 degrees and 35 degrees is the same as the difference between 75 degrees and 76 degrees No true zero even at 0 degrees there is still a temperature it still exists Ratio permits rank ordering of events with assumption of equal intervals and a true zero The difference between 10 mph and 20 mph is the same as the difference between 15 mph and 25 mph 10 mph There is a true zero if you are going 0 mph then nothing is happening o Be able to identify when each type of scale is being used o How do you identify whether a continuous variable uses an interval or ratio scale In an interval scale there is no true zero such as temperature Even if it is 0 degrees there is still a temperature So you can t say things such as twice as hot or half as cold In a ratio scale there is a true zero so 30 mph is HALF as much as 60 mph and 0 mph means no movement Know the three types of reliability and when they are most relevant E g when should you look for internal reliability vs when should you look for inter rater reliability Understand the various types of validity Test retest reliability Test is given twice and the scores are compared the greater the similarity between the scores the higher the reliability Interrater Reliabity 2 or more independent observers should come up with very similar findings Useful for naturalistic observations because often one observer may miss something or misinterpret a behavior Internal Reliability Answers are consistent no matter how the question is phrased o How can a scatterplot be used to evaluate reliability Correlation coefficient r respresents the degree of correlation If both the x and y variables are interval or ratio and the data appears to be linear then the data is reliable o Are correlations related to reliability In internal reliability items like I would change nothing in my life and there are few things in my life that I would want to change are correlated If the participant agrees to both of those items it would be reliable Chronbach s alpha data from a large participant pool is collected that asks the same question worded many different ways Chronbach s alpha is based on the average of correlations and number of terms between 0 1 Higher number means greater internal reliability Face Vailidity Is this a plausible measure of the variable in question is life satisfaction a measure of happiness Is head circumference a measure of intelligence Content Validity does it capture all parts of a defined construct If you are measuring intelligence to have good content validity you would have to address ability to reason planning abstract thinking complex comprehension etc Criterion Validity How well does a measure relate to a specific outcome Predictive how well does the measure predict the outcome Concurrent how well does the measure correlate with the outcome o What is the difference between content and construct validity Content Validity is the estimate of how much a measure represents every single element of a construct Construct Validity defines how well a test or experiment measures up to its claims o What is the usefulness of Known Groups Evidence for Predictive Concurrent Validity Known groups paradigm do scores on a measure discriminate among a set of groups whose behavior is already well understood This is useful for predictive validity because if the test accurately predicts something that is already known to be true then it is valid If the test has good concurrent validity then it will accurately correlate with the outcome discriminant validity between two variables o How can someone determine whether there is convergent or Convergent Validity measure should correlate strongly with other measures of the same construct If a scale is a good measure of depression it should correlate well with other measures of depression Discriminant Validity measure should not correlate well with measures of other distinct constructs There should be no correlation with measures of other traits Measure of depression should not correlate with measures of ice cream preference Chapter 6 Surveys and Observations Describing What People Do Recognize the strengths weaknesses associated with different question formats on surveys E g open ended forced choice etc Yes no questions allow little explanation Forced Choice questions agree or disagree no opinion or not sure are not options Often there is a set of statements and you must put a check mark next to those that you agree with Multiple Choice questions Must select most suitable response Open ended questions Respondent must construct own answer Information is more complete but the participant may not understand what you are looking for some answers are skipped over and respondent may not think of enough information The individual answers a question by selecting a response alternative from a designated scale such as 5 strongly agree 4 agree 3 neutral 1 strongly disagree
View Full Document
Unlocking...