1 What are the characteristics of descriptive and quasi experimental designs Key Points for Exam 2 i Observe events of phenomenon to detail or categorize them or chart their 1 How do students interest in and perception of importance in math and sports change across schooling a Descriptive course b Quasi experimental i 2 kinds 1 Find naturally occurring groups and see if they differ a Quasi because you don t manipulate it i Gender before after something happens 2 Compare an experimental group to a control group WITHOUT randomly assigning a You still manipulate independent variable but no random assignment to conditions i Pre existing classrooms 1 Cannot claim that groups were the same to begin with ii Are men or women more likely to be depressed 1 Gender was not something that was manipulated 2 Naturally occurring groups a Just looking to see if the groups differ 2 What are the scales of measurement we ve discussed What kinds of statistics are appropriate for each scale a Nominal Scale i Categories have different names but order doesn t matter ii Reflects qualitative differences 1 Sex ethnicity sexual orientation iii Can be assigned numbers but they don t mean anything 1 Male 1 and female 2 iv Statistical procedures 1 Frequency counts mode Chi Square v Do more men or women go into Psychology 1 Nominal scale gender 2 Nominal scale academic major a You can t compute an average major you can say that two individuals differ in gender or major b Ordinal Scale i Categories have different names AND are organized sequentially 1 A horse race ii We can determine If there are 1 2 The direction differences of the differences iii We CANNOT say how big those differences are iv Appropriate Statistics 1 Percentiles correlation Mann Whitney U v Example 1 Taste Test between 3 different Colas a Pepsi b Coke c Publix Brand 2 Given this ranking I can t tell how much more the person liked Pepsi 3 compared to Coke There is no numerical value to that difference Likert scales are also considered ordinal but a special case can treat as interval a Is the distance between strongly disagree and disagree the same as strongly agree and agree Pretty objective c Interval Scale i Categories are organized sequentially and all categories are the same size 1 What time is it Intervals ii Has an arbitrary zero point 1 Temperature in Celsius we chose to put 0 as freezing point doesn t mean there is no temperature iii Can determine how large or small differences are 1 Difference between 10 and 20 degrees is the same as 50 degrees to 60 degrees iv Statistical Procedures d Ratio Scale 1 Mean median mode standard deviation correlation t test ANOVA i Consists of equal ordered categories anchored by a non arbitrary zero point If you got a zero on a test you actually got zero right none 1 2 Zero is telling us information it is nothing ii Zero is the complete absence of the variable being measured iii Appropriate Statistics 1 Mean median mode standard deviation correlation t test ANOVA 3 What are some factors you need to consider when choosing a measure ratios a Information Yielded i A nominal scale yields the least information ii An ordinal scale adds some crude information iii Interval and ratio scales yield the most information b Statistical Tests Available i The statistical tests available for nominal and ordinal data are less powerful than those available for interval and ratio data ii Use the scale that allows you to use the most powerful statistical test c Ecological Validity i Sometimes your research requires you to use a measure that reflects what people do in the real world a nominal guilty not guilty verdict ii You want one with the most info and statistical options BUT that also reflects your real life construct the best iii Balance between these d Sensitivity i Is a measure sensitive enough to detect behavior change 1 Marital happiness a Divorce or of arguments per month 4 What types of dependent measures are available self report behavioral physiological and implicit Know about each of these a Behavioral Measure i Record actual behavior of subjects 1 Test assessment 2 Observation of some behavior ii You can record different things 1 Frequency number of behaviors that occur Latency Amount of time it takes for a behavior to occur 2 3 Number of errors number of incorrect responses made iii Cannot provide answers regarding causes of behavior b Naturalistic Observation i Unobtrusive observations of subjects naturally occurring behavior are made 1 Shopping behavior in malls play behavior in day care chimps in the wild ii Observer bias iii Reactivity 1 You see what you want to see c Physiological Measure behavior i Physical measure of body function ii Typically requires special equipment iii Most are noninvasive 1 Once people notice they are being observed they may change their 1 Some require an invasive procedure implanting electrode on brain of rat d Self Report Measure iv Allow you to make precise measurements of arousal of a subject s body 1 Must infer psychological states i Participants report on their own behavior or state of mind ii Self report measures are popular and easy to use but may have questionable 1 You cannot be sure that a participant is telling the truth when using self reliability and validity e Implicit Measures report Implicit Associations Test IAT is a popular example i ii When you are trying to ask people things that will make them feel bad if they iii answer honestly you use this JOY if you are stereotyped against African Americans you are more likely to put the word with European American 5 What does it mean for a measure to be reliable What are the different types of reliability a A reliable measure produces similar results when repeated measurements are made under identical conditions i Want to hit the bullseye consistently b You want a measure that does not have a large amount of error c Correlations with itself d Types of Reliability i Test Retest 1 Successive measurements 2 Score at Time 1 Score at Time 2 Correlation 3 Best for stable characteristics ii Internal 1 Split Half Internal Consistency a b Combine two forms and administer together c Form A Form B Combined Test Correlation 2 Item Level a Looks at the correlation of each item with the score of the remaining items together Item total correlation Item 1 Combined score on items 2 3 4 Correlation Item 2 Combined score on items 1 3 4 Correlation b c d Cronbach s alpha Very Common i Usually want it to be above 0 7 or 0 8 iii Interrater
View Full Document
Unlocking...