Lecture 2: Scientific Methods Flashcards
What is a theory?
Explanation based on observations.
What is a hypothesis?
Prediction based on theory.
What is research?
Testing the prediction.
What do you do if your research supports your hypothesis?
Refine your theory and test new hypotheses.
What do you do if your research refutes your hypothesis?
Discard your theory or revise and retest new theory.
What are the three ways of gathering data?
1) Descriptive Research
2) Correlational Research
3) Experimental Research
What is Descriptive Research?
- enables detailed description of behaviour.
- behaviour is not controlled
- cannot explain WHY behaviour occurred
- e.g. case-study, observation, self-report.
What is Correlational Research?
- allows us to answer questions otherwise unethical to investigate experimentally.
- look at natural relationship between variables without intervention.
What is Experimental Research?
- allows us to examine effect of IV on DV.
- If experiment is done well, cause and effect can be established.
What are the pros and cons of case studies?
DESCRIPTIVE:
PROS:
1) low in generalisability.
2) useful starting points.
CONS:
1) Correlation vs. Causations: association between two things does not imply cause and effect relationship.
2) Inability to verify some or all information put forward by single individual.
- can it be applied to a wider population?
- deliberately false answers/information given.
3) concerns need to be raised regarding inference of cause and effect relationships.
What are the pros and cons of naturalistic observation?
DESCRIPTIVE
PROS:
1) High in external validity (same as in real life).
CONS:
1) Low in internal validity
(depends on reactivity of observed).
What is the Hawthorne effect?
- Act of observation changes behaviour.
- Reduced external validity of phenomena being studied.
- Reducing one’s presence maintains naturalism BUT involves ethical issues.
What is an Observer Bias?
Having preconceptions regarding behaviour which influences observation of behaviour (e.g. societal/cultural norms).
What is the Experimenter Expectancy Effect?
When the experimenter has expectations as to what they they will observe (e.g. rat observation).
What are Experimenter Effects?
Experimenter has biased investment in one condition over another.
What is Covert Observation?
identity of the researcher, the nature of the research project, and the fact that participants are being observed are concealed from those who are being studied.
- might reduce impact of experimenter effects.
- might not reduce impact of expectancy effects.
What are the two classic expectancy effects?
1) Placebo effect: expectations are positive.
2) Nocebo effect: expectations are negative.
What are the pros and cons of self-report/surveys?
DESCRIPTIVE
PROS:
1) easy to administer
2) cost-effective
3) relatively fast way to collect lots of data
4) contrast limited generalisability of case-study/observation
CONS:
1) questions can be poorly worded.
2) answer boxes that do not allow a full range/representation of answers.
3) possible to lie.
What is important to consider when taking a survey/self-report?
- Where data came from.
- Other variables that effect results.
What are the aims of correlational research?
1) to specify relationship between two variables.
2) to be able to predict one from the other.
What are two factors to consider when using correlational research?
You must establish the DIRECTION and SIZE of 2 variables.
What is direction in correlational research?
Positive: as x increases y increases (same for decrease).
Negative: as x increases y decreases (same for reverse).
What is size in correlational research?
-How well one thing can be predicted from another.
Value +1 or -1: perfect correlation between x and y.
Value 0: no relationship between x and y.
What are the pros and cons of correlational research?
PROS:
1) we need correlational design because there are some things we cannot ethically control.
CONS:
1) An association between two things does not imply a cause and effect relationship.
- CORRELATIONS DESCRIBE RELATIONS RATHER THAN CAUSATIONS.
2) Third variable problem: 3rd factor which drives relationship between x and y.
3) Directionality problem: does x effect y or y effect x?
What is the aim of the experimental method?
to establish causation by controlling the environment.
What are the different types of variables?
1) Independent: the thing you change
2) Dependent: the thing you measure
3) Extraneous: other things we must control
What is a solution to an experiment in which we cannot control everything?
Random allocation of individual to group.
- does not entirely eliminate extraneous variables
- does help reduces chances of effect being due to EVs.
What are the two fundamental characteristics of experiments?
1) Manipulation of IV.
2) Random assignment of participants to groups.
What is an operational definition?
Establishes what you mean by the fact you are measuring.
-helps overcome conceptual vagueness.
What is reliability?
refers to the consistency of measurement.
What is test-retest reliability?
consistency of measurement across time (same results over time).
What is inter-rater reliability?
consistency of measurement across people.
What are two things to be wary of in experiments?
Random error: human error.
Systematic error: faulty equipment.
What is validity?
refers to the meaningfulness of measurement.
-e.g. a ruler is a reliable but invalid way to measure intelligence.
What is convergent validity?
Suggest that multiple measures that tap the same thing should yield similar results.
E.g. more smiles, jobs, donations = treated better.
What are the differences between lab-based studies and field-based studies?
Lab-based studies:
- high internal validity
- low external validity
Field-based studies:
- low internal validity
- high external validity
What are descriptive statistics?
A summary of what the data shows after collecting data from a study.
Traditionally, what 2 pieces of information do descriptive statistics contain?
1) measures of central tendency (average)
- mean
- mode
- median
2) measures of variance
- range
- standard deviation
What are the two types of privacy in the ethics of psychology?
1) Confidentiality: collecting personal information and not sharing it.
2) Anonymity: personal information not collected.
What is a participation risk?
Ethics board weigh risk/benefit ration to see whether research is important enough.
