Study Guides (238,467)
Canada (115,151)
York University (9,811)
Sociology (180)
SOCI 2030 (26)

Final Exam Notes.docx

9 Pages
Unlock Document

York University
SOCI 2030
Tracy Supruniuk

Lecture July 12’12 EXPERIMENTAL RESEARCH Appropriate Research Questions - Experiments are best research design for establishing causality - Causality: variable A causes change for variable B - Experiments are best for establishing causality because it’s the easiest to control - Experiments allow for the greatest control of the 3 criteria of causality: 1. First establish an association between the variables; variables need to be associated in some way 2. You need to establish a temporal order one has to happen before the other 3. You need to rule out rival causal factor - Studies best suited for experimental research are those you can best manipulate the independent variable - Experimental research not suited for studies where you can’t manipulate the independent variable - Experimental research better for looking at small set variable - Better for studying relatively small groups Terms Used in Experiment 1. Experimental group/treatment - Experimental groups- experiences the modification or stimulus - In an experiment there is 2 groups: 1 experimental & 1 control group - Whatever modification your giving to the experimental group is called the treatment 2. Control Group - Doesn’t receive modification, stimulus, or treatment - Served as comparison group - Usually the case they don’t know about their other group - Some level of deception is required in deception due to Hawthorne effect - Whenever you do an experiment you need to measure dependant variable 3. Pre-test - When you measure the dependant variable before it’s called pre-test 4. Post-test - You measure after the treatment is given  When doing experimental/control group experiment you want to do pre+post test  When doing an experimental/control group research you want to randomly select people  Goal: both groups to be similar you want to have similar characteristics 5. Matching - You randomly select people based on similar characteristics into each group is called matching - E.g. 100 people for drug test: 15 men + 15 women for control group & 15 men + 15 women for experimental group Different Experimental Designs - 3 general types 1. True Experiments - Has 3 general characteristics a) Involve random assignment- groups are randomly selected b) Experimental/control group c) Involved pre/post test - 3 general types of true experiments i) Class Experiment E= equivalence; groups are randomly selected O= observations X= treatment # by O= time observation has been done E O₁ X O₂- got treatment (EXPERIMENTAL GROUP) E O₁ O₂- CONTROL GROUP - Each line represents a different group - Observed same time - Pre-test - Post-test - One of the best designs for establishing causality design ii) Post-test only Control Group Design - On a post test - Used when no bias result - Same as classical but no pre-test E X O₂ E O₂ - Difficult to rule out rival causal factors (disadvantages) iii) Solomon Four Group Design - Combines both groups to find out strength and weaknesses E O₁ X O₂ E O₁ O₂ E X O₂ E O₂ 2. Pre-experimental Designs - Don’t use random assignment - 3 types of pre-experimental designs - Researchers avoid using this type unless exploratory a) 1 Shot Case Study - 1 group and 1 treatment - And NO pretest XO₂ b) 1 Group Pre/Post test Design - Same as 1 shot case study design but has post test O₁XO₂ c) Static Group Comparison - You add a control group but no pretest XO₂ O₂ 3. Quasi Experimental Designs (NON RANDOM) - 3 general types a) Time series design O₁ O₂ O₃ (pre-test) X O₄ O₅ O₆ b) Multiple times series design - Same as time series design but looks at more than 1 group - The other groups will be experimental and control groups O₁ O₂ O₃ x O₄ O₅ O₆ O₁ O₂ O₃ O₄ O₅ O₆ c) Non-Equivalent Control Quasi Design - 2 groups involved - More than 1 experimental group - Group1- O₁ X O₂ - Group 2- O₁ X O₂ - Control group- O₁ O₂ Different Issues Involved with the Experiment - You need to be concerned with internal and external validity - Internal validity- change in dependant variable - External validity- refers to your ability to generalize results in your study Issues to Related to Internal Validity A) Selection Bias - Examining not equivalent groups (not randomly assigned) - Randomly assigns groups B) History - When some outside event unrelated to your experiment and it effects your outside variable - Outside event less of an issue if it’s both groups E.g. Rainy day- causes depression C) Local History - Only effects 1 of the outside group D) Maturation - Change of dependant variable is due to age on time process not the treatments - Happens to longitudinal studies E) Experimental Morality - Participants stop participating or die or move away F) Testing - Test effect G) Instrumentation - Change in how a concept is measured during course of study - E.g. Scale H) Diffusion of Treatment - Treatment is your experimental groups moves 2 your control groups and contaminates in it I) Compensatory Behavior - They might compensate to not have the treatment (control group) J) Experimenter Effects - To remain objective in study the do things consciously to effect study - E.g. Body language- gives clues to prevent experimenter effect you do double blind study ----> Researcher doesn`t know whose getting treatment and whose getting placebo Lecture July 17`12 SURVEY CONSTRUCTION Appropriate Topics for Survey Research - Good for social research - Describes a respondents view, explores an issue, evaluates a program (e.g. customer satisfaction survey) - Surveys can be used to explain a phenomenon - Surveys can measure many variables at the same time which you can rule out rival factors - Opinion studies are used to assess people views of social issues - Surveys can look at peoples self reported behavior (e.g. do you smoke marijuana) - Better for people to report on recent behavior instead of past because its easier to remember - When ask about past behavior is called recollective questions (e.g. did you work during school) you ask general questions - Surveys are good for asking people about their specific characteristics also referred to demographic
More Less

Related notes for SOCI 2030

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.