PSYCO104 Lecture 5: Spet 20. 2016

• School: University of Alberta
• Department: Psychology
• Course: PSYCO104
• Professor: Blaine Mullins

Document Summary

You need a random assignment and a control/manipulated variable. You need both experimental (receive treatment; iv) and controlled (not receiving anything, no. Correlational design: measuring how close 2 or more variables are related: correlation coef cient. Strength (how closely the dots are together, as x goes up y also goes up) Direction: use correlation cuz theres things that you cant control. Ex. gender, religious belief cannot be measured into an experiment. Why do use correlation: if you cannot manipulate variable. Upside: have more info than just saying its different you can say how its related and whether if its positive or negative. Naturalistic observation: go into the real world and observe and relate it to the rest of the population: gives opportunity to observe real behaviour, downside: can"t really control variables, time consuming, expensive. Any design that has some elements of the experimental design but is not a true experiment: cant completely say its an experiment.