research designsX

Views:
 
Category: Education
     
 

Presentation Description

Research designs for the social science surveys and experiments: one sample, separate groups, repeated measures, correlational.

Comments

Presentation Transcript

Research designs : 

Research designs headlessprofessor

Objective : 

Objective Describe the logic of science

Purpose of Science : 

Purpose of Science Understand

Purpose of Science : 

Purpose of Science Understand Predict

Purpose of Science : 

Purpose of Science Understand Predict Control

Logic of science : 

Logic of science

Example : 

Example Marriage helps increase your happiness.

Hypothesis : 

Hypothesis A prediction as to the results you expect to find

Hypothesis = a prediction which guides research : 

Hypothesis = a prediction which guides research (hypotheses are generated by theories, but are testable by data)

Population = entire class of subjects studied : 

Population = entire class of subjects studied

Sample = those subjects actually observed : 

Sample = those subjects actually observed Should be large and representative of the population

Group : 

Group A subsample

Group : 

Group A subsample Determined by background factors

Group : 

Group A subsample Determined by background factors Assigned in experiments

Research Design #1 : 

Research Design #1 Sample vs. Norms

Representative Sample : 

Representative Sample One that does not over-represent or under-represent any particular segment within the population

Representative Sample : 

Representative Sample Does not over-represent or under-represent any particular segment within the population The proportion of sub-samples (groups) in the sample should match the proportion of these segments of the population

Hypothesis : 

Hypothesis Married people should be happier than average.

Confounding variables : 

Confounding variables Anything leading to sampling bias

Confounding variables : 

Confounding variables Anything leading to sampling bias Geography

Confounding variables : 

Confounding variables Anything leading to sampling bias Geography Elapsed time

Research Design #2 : 

Research Design #2 Separate groups

Experiments manipulate an independent variable : 

Experiments manipulate an independent variable Separate groups (subjects are randomly assigned and then treated differently): between subjects comparison on DV

Separate groups design : 

Separate groups design

A placebo helps control patient expectations : 

A placebo helps control patient expectations New med is given to experimental group Placebo is given to control group Pills taste the same Patients do not know which is which Researchers measuring DV do not know until after data are collected

Hypothesis : 

Hypothesis Married people should be happier than unmarried people.

Confounding variables : 

Confounding variables Anything influencing grouping

Confounding variables : 

Confounding variables Maybe happy people are more likely to find spouses

Confounding variables : 

Confounding variables Maybe happy people are less likely to get divorced

Research Design #3 : 

Research Design #3 Repeated measures

Experiments manipulate an independent variable : 

Experiments manipulate an independent variable Separate groups (subjects are randomly assigned and then treated differently): between subjects comparison on DV Repeated measures (entire sample is measured on DV before and after the treatment): within subjects comparison

Repeated measures design : 

Repeated measures design

Problems with repeated measures experiments : 

Problems with repeated measures experiments Fatigue & boredom effect Practice effect Natural course of a disorder Attrition of subjects changes the sample

Hypothesis : 

Hypothesis People should be happier after marriage than before.

Confounding variables : 

Confounding variables Maturation

Confounding variables : 

Confounding variables Maturation Change in wealth

Confounding variables : 

Confounding variables Maturation Change in wealth Historical periods

Research Design #4 : 

Research Design #4 correlational

Hypothesis : 

Hypothesis People who have been married longer should be happier.

Confounding variables : 

Confounding variables Age

Confounding variables : 

Confounding variables Age Cohort

Research designs : 

Research designs headlessprofessor