This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Research Methods Research Methods Some definitions….. Some definitions….. Theory: A broad explanation that explains all of the current observations A testable prediction specific Very specific descriptions of how a hyp will be tested and how the variables will be manipulated or measured Hypothesis: Operational definition: An observation Prior research Curiosity or insight A QUESTION Formulate an explanation: Specify a theory and develop a hypothesis Carry out research Purposes of Research Purposes of Research
1. To describe what are different styles of discipline? how does divorce affect children? when do children begin to read? is discipline style related to school performance? is divorce associated with depression in children? is how much time parents spend reading related to children’s reading ability? 2. To predict Purposes of Research, cont Purposes of Research, cont
1. Explanation does an authoritarian style of parenting cause children to do better in school? does parental divorce cause depression in children? does being read to increase the ability to read ? Descriptive methods Descriptive methods
1. Archival research: Strengths: Inexpensive; helpful in starting a new research topic Data may be incomplete; unusual events more likely to be recorded; researcher has no control Weaknesses: Example: descriptions of children’s behavior in old diaries; changes in SAT scores 1.. 1 Naturalistic observation/ structured observation Strengths: Weaknesses Avoids artificiality Researcher has no control; bias if people know they are being watched 2. Case studies Example: children on playgrounds; apes in wild Strengths Weaknesses a lot of detail and depth; useful for studying rare events Cases may not be similar to population at large Example: prodigies, certain disabilities 1.. 1 Surveys (selfreports) Strengths: Can gather a large amount of data Question wording; social desirablitiy; sampling problems; memory problems Weaknesses: Example: interviews, questionnaires Correlations Correlations A measure of how strongly two variables are related to one another Range is from 0 to 1 Indicates the strength Can be + or Indicates the direction Size and direction are interpreted separately Interpreting correlations Interpreting correlations Number (01): size; bigger numbers = stronger corr Sign (+ or ): direction +: both variables change in the same direction : variables change in the opposite direction students who spend more time studying tend to get higher grades students who spend more time in bars tend to get lower grades Related, but not one causes the other Any correlation could be interpreted three ways: Correlation does not imply Correlation does not imply causation A causes B (being read to causes children to be better readers) B causes A (children who are better readers cause parents to read to them more) C causes both A and B (some unmeasured variable leads to both higher grades and more studying) Parental value in reading, parental involvement overall, parental education, parental intelligence How can we get at cause then? How can we get at cause then? Experiments! Independent variable research design in which the researcher manipulates one variable while controlling all other conditions the variable the researcher manipulates the measured outcome in a study Dependent variable Hypothesis: The more time parents spend reading to their children, the better readers the children will become. Experimental group Control group Read to their children every evening for 30 minutes After 6 months, child’s reading is tested Read to their children every evening for 10 minutes After 6 months, child’s reading is tested Operationalization Independent variable: Dependent variable: Other variables (what needs to be held constant) Random Assignment Random Assignment Is a necessary component of an experiment
Equalizes groups Researcher can say the IV is the cause Why do correlational research? Why do correlational research? Goal is to understand causes of behavior, mental processes, but…. Sometimes experiments are impossible Unethical No independent variable Remember: Nonexperimental methods ARE scientific methods Artificiality Person variables Research Designs Research Designs Crosssectional Participants (Ps) are studied one time Ps are of varying ages Can’t really say that differences are due to time or change Possible cohort effect Longitudinal Ps are studied over a period of time Ps are same age at start of study Can say that changes are due to time or age Still a possible cohort effect, but a different kind Research Designs, cont. Research Designs, cont. Hybrid designs Also called sequential or crosslag Combines crosssectional and longitudinal Hybrid Design Hybrid Design
Time 1: 8 20 Time 2: 10 Time 3: 12 Time 4: 14 Time 5: 16 Time 6: 18 12 14 16 18 20 18 20 16 ...
View Full Document
This note was uploaded on 12/21/2010 for the course PSYCH 4 taught by Professor Ritabutterfield during the Fall '10 term at Santa Rosa.
- Fall '10