Strategies

• Descriptive research measures variables as they exist. Doesn’t explain relationships between them. Only reports on variables themselves to define them
• What is the typical number of text messages received a day?

  Descriptive research
• What is the average number of hours slept a day?

  Descriptive research
• What % of students voted in presidential election?

  Descriptive research
• Descriptive research typically uses descriptive statistics: percents, mean, range.
• Correlation is a measure of relationship between variables. Doesn’t explain the relationship or cause and effect, only examines strength and direction of relationship
• Is there a relationship between sunlight and depression?

  Correlation
• Is there an association between exercise and depression?

  Correlation
• Is there a link between genetics and depression?

  Correlation
• Time of sunlight and levels of depression?

  Pearson’s r (correlation)
• Time of sunlight and having depression (yes/no)?

  Point biserial (one continuous and one dichotomous)
• Sunlight exposure (yes/no) and level of depression?

  Point biserial
• Sunlight exposure (yes/no) and having depression (yes/no)?

  Phi (two dichotomous)
• Point biserial is one dichotomous (yes/no) and one continuous (miles)
• two dichotomous variables?

  Phi
• Experimental research: cause an effect relationship by random assignment, manipulating one or more variables to see how it affects the dependent variable.
• Testing the effect of exercise v. no exercise on cholesterol?

  Experimental
• Testing effect of studying with music v. without music on test performance?

  Experimental
• Testing effect of caffeine v. no caffeine on heart rate?

  Experimental
• Wuasi-experimental: cause and effect relationship and limits threats to internal validity but lacks random assignment
• Without random assignment, administer smoking cessation therapy to one group of smokers and not another group, then compare cessation rates?

  Quasi-experimental
• Without random assignment, implement an after school program in one school and not another school then compare grades?

  Quasi-experimental
• difference between experimental and quasi is that quasi doesn’t have random assignment
• nonexperimental research determines if relationship exists by comparing different groups of scores but it doesn’t attempt to minimize internal validity threats or explain the relationship
• Is there a difference in verbal skills between men and women?


  Nonexperimental
• Is there a different in math skills between left and right-handers?

  Nonexperimental
• Is there a difference in long term memory between cats and dogs?


  Nonexperimental
• Nonexperimental and correlational research have same goal: to determine if a relationship exists and both do not infer causation
• A difference between groups in nonexperimental design indicates a relationship between variables
• nonexperimental design doesn’t try to explain a relationship, just like correlational and descriptive
• descriptive research examines individual variables instead of relationships between variables
• correlational research examines for relationships between variables by measuring two or more variables for each participant
• experimental, quasi-experimental, and nonexperimental examine for relationships between variables by comparing two or more groups of scores
• between subjects design: participants are randomly assigned to a condition of the experiment and are not exposed to the other conditions (participants only receive one condition)
• within subjects design: participants are exposed to multiple conditions or all conditions at different times
• comparing SAT math scores between physics and chemistry majors with an independent samples t test?

  Between subjects design
• comparing SAT math and reading scores in biology majors with a dependent samples t test?

  Within samples design
• Data that you analyze with a one-way ANOVA can be analyzed with independent-samples t test
• One-way between subjects ANOVA test the effect of drug A on one group’s symptoms and deug B on another group’s symptoms
• one way repeated measures ANOVA test the effect of drug A and B on one group’s symptoms