Analyzing Results of Data • If descriptive research (nominal data) compare group percentages → • If experimental or quasi-experimental and have interval or ratio data (DVs) com → - pare groups means via inferential statistics • If continuous predictor variables (non-experiemental) correlation → Descriptive Stats • Central tendency ◦ What is the whole sample like ◦ Mean (m), median, mode (e.g., if nominal scale used) • Variability ◦ What does the distribution look like ◦ Range: highest to lowest ◦ Variance (s2): measure of dispersion (spread) ◦ Standard deviation (SD): average distance from the mean (square root of variance) Goals of Research • Description, Prediction , Causation, Explanation, Application Relationship Between Variables • For example: ◦ Height and winners of presidential elections ◦ Hurricane exposure and rate of PTSD • Are two variables related ◦ AKA, do they vary together? • Correlation ◦ Change in one indicates change in the other Simple Correlation • Degree of relationship based on scores of individuals ◦ E.g., SAT scores and college grades ◦ Just need a score on the 2 variables for each participant in the sample • Correlation coefficient ◦ Statistic used to indicate the degree to which 2 variables are related to one another in a linear fashion • Pearson's r ◦ Sign refers to direction (+ or -) ◦ Magnitude refers to strengths (-1.00 to +1.00) • Positive correlation ◦ As you increase one variable the other variables increases as well ◦ The variables go up and down together ◦ Ex., time exercising and calories burned • Negative correlation ◦ Inverse relationship ◦ As you increase one, you decrease the other ◦ Move in the opposite direction ◦ Ex., number of kids and hours of sleep Significance and Strength • Statistical significance
• Related to the strength of a relationship • Significant correlation ◦ Sample of 100 participants with a Pierson's r of .4, but does it relate to the population ◦ Unlikely that in the parent population that the relationship is zero • r is significant when there is a low probability of the relationship being zero in the parent population ◦ Bigger r, larger sample, less variance • Strength (effect size) ◦ < .20 are weak ■
You've reached the end of your free preview.
Want to read all 5 pages?