Test 3 Part 2 Flashcards
Benefits of factorial designs and why we learn to interpret interaction graphs
- Two “main effects” in one study is like having two studies with one IV each
- Economy of participants (get more out of participants)
- Interactions: can tell us whether the effect of one variable stays the same over time or changes depending on levels of the other variable
Why we learn
- To understand psychological phenomena, we need to examine interactions
- Learning to visualize data can be difficult at first, but once u get it it makes interpreting data much easier
- Easier to read journal articles
Nominal, ordinal, interval, and ratio
- Nominal: categories or groups
•ex: ethnicity, levels of IV - Ordinal: rankings- uneven spaces between scores (much more rare)
•ex: olympic metals, piaget developmental stages - Interval: even spaces between scores, 0 is arbitrary
•ex: degrees, IQ - Ratio: even spaces between scores with a true 0
•ex: height, weight, reaction time
Measures of central tendency
- Mean: add the scores and divide by N
•most affected by outliers (ex: income) but is used most often because
*with increasing sample size, a single extreme score has less effect on the mean, maximizes use of all the data, and has mathematical properties useful for statistical analysis - Median: the score that divides the group in half (50% below and 50% above)
- Mode: most occurring number
How much people scores differ from these central tendencies is determined by
1. Visual: frequency of distribution of scores
•looking at graphs
•histograms have no gaps (different than a bar graph) to show that its a continuos variable
- Range: maximum score minus minimum score
- Standard deviation (SD) and variance (s2 - s squared): how much variability in scores is around the mean, how stable is this mean, and how representative is this mean of any given score
•SD: think about it as an index of how far each score is from the average mean, standardized means we can learn to interpret it, tells us what is in the typical range
•s2: sum of squared deviations around mean divided by N-1
ex: Wechsler intelligence scales: average is 100, and 1 SD above it is 115 and 1 SD below it is 85, making the SD 15
Descriptive statistics
ASK THE GIRLS
Can help us understand our participants by expressing a correlation as a percentage
•central tendencies summarize how participants scored, and variability summarizes how widely the distribution of scores was spread
•r2: a proportion
•ex: r2 of .31 means that 31% of the variability in expected grade is shared with self reported GPA from las year (and 69% is not)
•shared variance
circles on the page