PSYCH 2300 Lecture Notes - Lecture 8: Penguin Group, Convergent Validity, Statistical Significance
Document Summary
Psy 2300: chapter 8 - bivariate correlational research. Closer to 1 or -1 indicates strength. Sign indicates association (pos+, neg-, or 0) Use a bar graph if you have at least 1 categorical variable. 2 measured variables in a study make it correlational. Can estimate effect sizes by looking at scatterplots. Stronger correlations/effect sizes = more accurate predictions. P-value: probability of finding the observed effect size given that it comes from a pop. in which the association is actually 0. Larger sample sizes create smaller r values to declare statistical significance. Is not really accurate, does not equate practical significance. We can just say it"s unlikely that we got these results if the hypothesis is true. In smaller samples, outliers can have a bigger effect on r. Not working with a full set of data can affect your results. Ex) too little of a drug won"t arouse you much, but too much of a drug can be toxic.