Chapter2 the methods_of_psychological_research

The Methods of Psychological Research

Chapter 2pp. 34-72

The Scientific Method

• How do we advance knowledge in our field of research?– Formulate the research question based on a careful review of the literature

– Generate hypotheses based on what is already known

– Design an experiment using a paradigm and methods appropriate to the research question, incorporating carefully designed controls

• How do we advance knowledge in our field of research?– Obtain ethics approval for the study

– Collect data with careful attention to random sampling and random assignment

– Analyze data and formulate interpretations using appropriate statistical tools

The Scientific Method

• How do we advance knowledge in our field of research?– Communicate results in the form of published articles and presentation at scientific conferences (peer review process)

– Replicate!– Generate the next research question based on new results

The Scientific Method

• What is a scientific hypothesis?– A proposed explanation for a phenomenon

– Based on previous observations or extensions of scientific theories

– A provisional idea whose merit is to be evaluated

– Must be testable and falsifiable– If confirmed, it is not “proven,” but remains provisional (i.e., for the time being)

The Scientific Method

How do we test hypotheses?• Various methods of gathering data• Descriptive

– To develop description of the phenomenon

• Correlational– Is there an association between two variables?

– DOES NOT IMPLY CAUSATION

• Experimental– Independent variable: manipulate– Dependent variable: measure

The Scientific Method

Steps in the Scientific Method

• Four main steps of scientific method1.Generate a Question2.Gather Information3.Formulate Hypothesis4.Test Hypothesis and Interpret Results

• Operational Definition

Descriptive Research

• Goal– Observe and describe behaviour– Not explain causes

• Challenges– Reactivity– External validity

Techniques of Descriptive Research

1. Naturalistic Observation

2. Case Studies

3. Surveys

4. Psychological tests

Naturalistic Observation

http://www.oxfam.org.uk/coolplanet/ontheline/explore/journey/france/educ.htm

Naturalistic Observation

http://www.ok4me2.net/category/science-and-technology/biology/page/5/

Naturalistic Observation

Case Studies: Focusing on the Individual

• One or a few individuals who are studied in great depth

• Potential Problems– External validity– Verification

Surveys: Describing Populations

• Gather a limited amount of information from many people

• Potential problems–Obtaining a representative sample of participants–Are respondents being truthful?–Are respondents’ answers accurate?

Sampling from a Population

• Population

• Sample

• Random Sampling– Procedure for selecting a representative subset of a target population

– Procedure guarantees that everyone in the population has an equal likelihood of being selected for the example

Psychological Tests: Assessing Individual Differences

• Tests mainly designed to measure individual differences– Achievement tests

– Aptitude tests

– Intelligence tests

– Personality tests

Correlation Research: Predicting Behaviour

• Correlation– A statistic that indicates whether two variables are related or vary together in a systematic way

• Correlation Coefficient– Vary from +1.00 (perfect positive relationship) to 0.00 (no relationship) to -1.00 (perfect negative relationship)

• Correlation coefficient used to summarize whether two measures vary together

• Positive Correlation: One measure goes up, the other tends to go up as well

• Negative Correlation: One measure goes up, the other tends to go down

• Zero Correlation: Knowing value of one measure does not allow you to predict value of the other measure

Correlation Research: Predicting Behaviour

Correlations and Causality

• Determining whether a relationship (correlation) exists between two variables is important in everyday life

• Correlations are useful for describing patterns of relationships between variables

• Limited in value when it comes to understanding why behaviours occur or co-occur

Positive and Negative Correlation

Third Variables

• Reason causality from a correlation is not possible due to presence of other factors

• Two variables can appear to be connected but connection could be due to some third variable

Experimental Research: Explaining Behaviour

• Experimenters actively manipulate environment in order to observe effect of behaviour– Example:

•Deliberately expose one group of children to violent show, other group to nonviolent show and observe behaviour of each group

• Differs from simply recording, observing behaviour

Independent and Dependent Variables

• Independent Variable– The aspect of the environment that is manipulated in an experiment. It must consist of at least two conditions.•Example: Violent show and nonviolent show

• Dependent Variable– The behaviour that is measured in an experiment•Example: Amount of aggressive behaviour after watching the show

• Hypothesis tested by observing effect of manipulating independent variable

Dependent and Independent Variables

IndependentVariable

DependentVariable

Experimental Control?

• Experimental and control group must be similar except with regard to independent variable

• There must be no confounding variables– Confounding variables: Uncontrolled variables that change along with the independent variable

• Random assignment: Each participant has an equal chance of ending up in any group or condition in the experiment– Helps ensure that overall, experimental and control groups are similar

The Scientific Method

Can tapes with subliminal messages help smokers quit?

Null Hypothesis (no effect): After listening to a self-help tape, participants will not quit smoking

Alternative Hypothesis: After listening to a self-help tape, participants will quit smoking

Self measures for a period of 6-months (data:self-report)

- # of cigarettes per week- Desire to smoke per week

The Scientific Method

2 groups

subliminal message no subliminal message

Dependent Variable: Measure – smoking/urges

Independent Variable: Manipulate – type of tape

Experimental Group vs. Control Group

Expectancies and Biases in Experimental Research

• Participants may guess what the researcher expects to find

• Single-blind study• Double-blind study

• Ways of reducing expectancy effects:– Give placebo to those in the control group

Generalizing Experimental Conclusions

• Must consider– Would participants have behaved the same way outside of the artificial laboratory situation?•Example: Would children in the television study behave the same way at home or school?

• External validity

Statistics

• Mathematical techniques for summarizing and analyzing data

• Two main kinds– Descriptive statistics mathematically describe data

– Inferential statistics help researchers decide•Whether data represents population•Whether differences between groups can be attributed to chance alone

Central Tendency

• Mean– Summarizes observations into a single representative number (arithmetic average)

• Median– Middle point in a set of scores

• Mode– Most frequently occurring score in a set of scores

Observational Study of Hand Washing

• Mean: (4+4+5+ 8+9)/5 = 6

• Mode = 4

• Median: 4, 4, 5, 8, 9– Median = 5– What happens if there is no middle number?

– 2, 3, 4, 6, 6, 7– Median = (4+6)/2 = 5

Variability

• Variability– How much the scores in a set differ from one another•Example: Two classes might have the same average exam score, but one set of scores might be much more spread out

• Standard Deviation– Average distance of scores from the mean

• Range– Difference between the lowest score and the highest score

Inferential Statistics: Interpreting Data

• Mathematical techniques that help researchers estimate the likelihood that recorded behaviours are –representative of a population or –the likelihood that differences or relationships between observations are due to chance

Inferential Statistics

• Based on laws of probability

• Example:– Does a gender difference in scores mean there’s a real gender difference in the population?

– Consider: What is the likelihood of getting a difference of a certain size or greater by chance alone?

• Researchers calculate probability that results could have happened by chance– If less than 5% chance > conclude the difference is real (statistically significant)

Treating Research Participants Ethically

• Overall goal– Respect rights, dignity of participants

• Major components of research with human participants– Informed consent– Debriefing– Confidentiality

• Ethical treatment of animal research subjects

Informed Consent

• Process of gaining permission and providing explanation of– Any risks (physical or emotional) or other factors that might affect decision to participate

– What procedures are involved

• Participants may stop without penalty

• May deceive about the true purpose, but not about factors that affect decision to participate

Debriefing and Confidentiality

• Debriefing– After the experimental session is over, telling participants more about the true purpose

– Explain any deception that took place

• Confidentiality– Researcher does not discuss personal information obtained from the research without permission

Ethics of Animal Research

• Benefits of animal research– Increases experimental control

•Enables research that can’t be conducted with human participants

• Ethical issues in animal research– Informed consent can’t be obtained– May involve injury or death to the animal

• Researchers who conduct animal research must– Care for animals properly

•Minimize pain and discomfort

Summary: Conceptual Problems

• What is the best way to observe behaviour?– Naturalistic observation, case studies, surveys and psychological tests allow researchers to observe behaviour

– Statistics are used to summarize and interpret findings

• How can we predict behaviour?– Correlation research allows researchers to predict future behaviour

Summary: Conceptual Problems

• How do we determine the causes of behaviour?– Experimental research allows researchers to determine causes of behaviour, if they are properly controlled

• How do we make sure that our research participants are treated ethically?– Informed consent, debriefing, and confidentiality help ensure that researchers treat participants with dignity

– Research with animals involves different ethical standards but humane treatment is required