Research Methods (Chapter 2) Flashcards
What are the steps in the Scientific Method?
1) Identify the question of interest
2) Gather data, and form the hypothesis
3) Test the hypothesis
4) Analyze data, draw tentative conclusions and report findings
5) Build a body of knowledge, ask further questions
What is a Hypothesis?
Specific prediction about phenomenon, often phrased as an if…then… statement
What is a Theory?
Set of formal statements to explain why/how events/phenomena are related.
It is more comprehensive than a hypothesis, and is formed when evidence accumulates that support the statements
May be altered with new evidence
What is Hindsight, and what are some of the problems inherent within it?
The concept that “Oh, we knew that all along”. Looking back, the answer is obvious.
Problems are that past events are interpreted differently with opposing explanations, by different people. How can we tell which explanation is correct?
What makes a good Theory?
Good theories are testable, and supported by evidence.
What is the Law of Parsimony?
If two theories explain the same phenomenon equally well, the simpler explanation is preferred
What are Variables?
Anything that varies within and between people.
Many variables are not observable, and can mean different things to different people.
Operational Definition
Defines the variables being tested in terms of specific procedures used to produce or measure it.
Or, “for the purposes of this experiment, we are assuming that variable x is understood as being _____”
What are four ways to measure variables?
1) Self-Report
2) Measure Overt Behaviour
3) Psychological Tests
4) Physiological Measures
What is a self report, and what are some problems with it?
Have people give reports on their own behaviour through interviews, questionnaires or surveys.
Accuracy depends on honesty, and can be affected by Social Desirability
What is Social Desirability?
When a test participant responds in ways that they think they should respond, in order to be socially acceptable, and not necessarily how they actually feel.
What are some ways to Measure Overt Behaviour?
Study reaction time
Coding: develop categories/labels for specific behaviours
Unobtrusive Measures: Minimizes social desiribility, but the mere fact that one is being observed may influence behaviour
Archival Measures: Use existing data
What are Psychological Tests?
Tests developed by psychologists to measure specific constructs and variables
What are some Physiological Measures, and the problems inherent?
Blood tests, heart rate, brain activity
Problem is, we don’t always know what those variables mean.
What are the Three Methods of Research?
1) Descriptive Research: Aims to describe behaviour
2) Correlational Research: Helps to predict behaviour
3) Experimental Research: Helps to predict and influence behaviour
What is a Case Study?
In-depth analysis on single group or event.
Collects data through observations, tests, interviews, etc.
Pros: Allows for in-depth info on rare phenomena, can lead to new research hypothesis
Cons: Cannot determine Cause/Effect, Low generalizability, can suffer from observer bias
(Descriptive Research)
What is Naturalistic Observation?
Passive observation of behaviour, without interference
Pros: Can yield in-depth information
Cons: Cannot determine Cause/Effect, observation may influence subject’s behaviour
Survey Population
All individuals within group of interest
Survey Sample
Subset of population. Must be representative of the whole population to retain validity
What is a Survey?
Questionnaire given to wide group of people
Pros: Efficient way to get large amounts of information
Cons: No Cause/Effect; relies on self-report; if not representative, the conclusions are inaccurate.
What are Random anStratified Sampling?
Random Sampling: Subjects chosen at random, each member of the population has equal chance of being selected
Stratified Sampling: Select sample after dividing up the population based on important characteristics (more representative)
Correlational Research
Measurements of associations between naturally occurring variables and events.
Variables are measured, but not manipulated
Correlation does NOT prove causation, it merely states that there IS a relationship, and may allow for predictions to be made
Simple Correlational Study
1) Measure variable x
2) Measure variable y
3) Using statistics, determine if x and y are related
What is the Correlational Coefficient?
A statistic that indicates the direction and strength of relationship between two variables
Positive: If variable x increases, so does variable y
Negative: If variable x increases, variable y decreases
Range: From -1.00 to +1.00
Experimental Research
Examines cause and effect through three components:
1) One (or more) variable that is manipulated or controlled
2) Researcher measures if manipulation influenced other variables
3) Extraneous variables that might influence results are controlled
Explain process of Experimental Research
1) Start with equivalent groups of participants
2) Treatment for each group is completely identical, except for variable of interest
3) Manipulate variable of Interest
4) Measure Group’s response
What are Independent and Dependent Variables in Experimental Research?
Independent Variable (IV): The variable being manipulated or controlled
Dependent Variable (DV): Variables expected to be influenced by the Independent Variable
What are Experimental and Control Groups?
Experimental Group: Group of participants who receive some level of IV; there must be a minimum of two levels of IV
Control Group: Does not receive IV, or received Level 0 of IV
Simple Experimental Designs
Between Groups: Each group has different (but matched) participants. Using random assignment can balance uncontrollable factors
Repeated Measures: Each participant exposed to al levels of IV (take part in both Experimental and Control Groups). Use of counterbalancing (the order in which participants are exposed to each level of IV varies) ensures that no condition has advantage of another
Validity
How well does the experiment actually test what it intends to?
Internal Validity
The degree that an experiment supports clear causal conclusions (does it actually work?)
Can be Impacted by:
1) Compounding Variables: unexpected/unaccounted for variables that influence result
2) Placebo Effect:
3) Experimenter Expectation: Researchers can unintentionally influence participants’ to respond in ways that are not consistent with the hypothesis (Double-blind procedure can help solve this)
External Validity
The degree that the results can be generalized to other populations (does it work elsewhere?)
Increased through repetition
Use of Meta-Analysis (statistical procedure combining results of many replicated and related studies) can help prove validity
Incomplete Disclosure and Deception
Highly controversial (Stanford prison experiment, or Milford’s shock-teaching)
Participants must be debriefed following experiment, or it violates the notion of informed consent.
Only permitted when no alternative is feasible, and benefits outweigh the risks
Has become less common in recent years
CPA Code of Ethics
Protect/Promote welfare of participant
Avoid doing harm
Not to carry out studies unless probably benefit is proportionally greater than risk
Provide informed consent
Ensure coercion was not used to acquire consent
Ensure privacy and confidentiality
