you must keep track of greatness – realize that material things are only a possible way to achieve greatness o Prof’s advice: the job that allows you to contribute to society in a real way [and thus contributes to greatness] often pays less – less ethical jobs will likely pay you more Chapter Four: Measurement and Data Analysis [Lecture 2/6, 2/11, 2/13] - Pop quiz concepts: o Criterion validity o Type I error o Meta-analysis o Sheldon - Reliability o Reliability: the extent that results are consistent across measures If there is less measurement error – then there is greater reliability
The more measurement error present – the more unreliable the finding Meaning that the measurement will be consistent across different sorts of scenarios There are many different forms of reliability o Parallel forms reliability How well two different forms of the same test correlate with each other Used to ensure that multiple forms of the same measure/test are consistently measuring things [are correlated] Ex: (1) two different forms of the same neuropsychology test o In neuropsychology cannot give the same test multiple times because of practice effects – thus use multiple (slightly varied) forms of the same test [like a memory test] (2) SAT of 2012 and 2013 o Split half reliability How well scores on one half of the test correlate with those on the other half Multiple ways to do this: first half/second half, odds/evens Problem: because there are multiple ways to split a test, there will be many different correlations o Internal consistency [Chronbach’s alpha – reliable in the .7-.9 range] [**look at 2/6 lecture figures] The average of correlations obtained using all possible split-halves o “split-half correlations” Average of items 1,3,4 correlate .87 with items 2,5,6 on a 6 item survey o The average of all of these split half correlations is the alpha
Indicates how well the items on a scale measure the same construct In SPSS can also get an indicator of how good each item is Can tell how well each item correlates to the total overall scores on the test/survey o Ex: in a test of anxiety, if one of the questions has a .1 correlation to the rest of the test – should reword/remove it o Test-retest reliability Correlate scores from the same test taken at two different time points Measure of stability Ex: (1) IQ (2) Personality (3) Mood We expect things like IQ and personality to be similar over time o Look at the correlation of people taking these tests at a certain time and again at a later date IQ: tends to have the highest test-retest reliability of any psychology measure [.95 correlation – almost perfect] Personality: .7-.8 range; 5 year correlations are still over .5 o Thus personality is relatively stable over time – but it does change Mood: we expect this to have low test-retest reliability o This is correlated, but unstable o Inter-rater reliability Correlate scores from two different raters
You've reached the end of your free preview.
Want to read all 48 pages?