Research Methods Test 1
Broad generalizations from specific observations (e.g. all swans are white)
Predict observations should the theory be correct (start with hypothesis then collect data)
Who thought of falsificationism?
Name the 6 components of the cyclical process
General principle, deduction, prediction, specific instances, observations and induction
What type of statistics are useful for testing hypotheses?
What is baynesian stats?
Utilizes prior probabilities
Name the essential concepts of ethics
stress and psychological harm
privacy and confidentiality
care of animals
costs and benefits
When carrying out a study that may cause harm, one must _____ from the participants
forewarn to allow withdrawal
What are some negatives of using deception (even if told afterwards)
Unintended lasting effect
May distrust research and psychologists
If data is anon, do people have to sign consent forms?
No, participation is consent
Debriefing must be
Thorough and complete
Give a chance to ask questions/give feedback
Special populations for consent in NZ include, and who speaks for them
Younger than 16
Cognitive deficits/mental disorder
What organisation is responsible for NZ ethical principles?
NZ psychological society
What can expose the truth on fraud?
Replications of the study
Can't measure constructs directly so have to measure them indirectly through variables (e.g. well being)
a measurement tool that consistently generates a similar empirical estimate
Correlation over time for the same individuals
Low reliability: psychometrically poor scale or phenomenon inherently unstable
Internal reliability is measured with and what does it measure?
The average level of inter correlation among all of the items
What is considered a unacceptable, acceptable and excellent cronbach alpha?
Do items relate to the overall construct
to what extent does the scale predict expected outcomes e.g. does grit scale predict success in a job/school?
To what extent does the scale measure the intended hypothetical construct e.g a test for depression should only test that, not anxiety and stress too
the extent to which the scale correlates with scales that assess something similar
scale does not correlate with scales that are expected to be unrelated e.g IQ and grit are said to be unrelated
Looking for an non significant correlation (not negative)
What is the highest order, most abstract type of validity?
Construct. repeated demonstrations that the scale represents the intended construct over numerous and various contexts
Numerical values that indicate membership of a group e.g. gender
Based on rankings
Only feasible with relatively small groups
Interval continuous variables
Lots of values obtained between a minimum and maximum e.g. multiple likert scales
Most common (most stats rely on assumption of normal dist)
Similar to ordinal and interval but has a true zero point
Pros and cons of self reporting?
You know yourself the best
Easy and efficient
Format of questions
Awareness/memory of events may be off
Name the 4 types of response scales and what type of data they produce
Binary Yes/No - Categorical
Fill in the blanks - Categorical (varied)
Multiple choice - Categorical
Likert - Interval
How many options do most Likert Scales have?
5-7 (odd number so you can have a neutral response)
When wording questionnaires, what should you avoid
Double barrel questions
Emotive language (loaded questions)
What are some pros of negatively worded questions in questionaires?
Can give balance and stop people from just selecting the same answer down the column
What are some pros and cons of digitally administered questionnaires?
Can do anywhere
Can edit easily
Can branch off - e.g. iwi
Can compile data e.g. excel
Problems- can't step back easily
Fonts can be hard to read
Computer can die
What is stratified sampling?
Divide population along dimensions to be sure to sample proportionately across these dimensions
What is cluster sampling?
Obtain data from preexisting groups/clusters
What are the non-probability types of sampling?
Convenience - readily available
Quota - Obtain appropriate percentages of different groups from readily available sources
Purposive - Select those who fit purpose
Snowball - enlisting peoples friends/family - good for rare groups e.g. surfers
Most commonly sampled population in psychology?
Western, educated, industrialized, rich and democratic
3 biases of sampling
unrepresentative (sampling frame doesn't represent pop.)
Self selection bias
What is ESM?
Event Sampling Method
Daily diary method e.g. app for mood over time
Pros and cons of ESM?
Catches phenomena close to the time
Multiple assessments increases validity
Can identify contexts
Cons- individuals may not want their day interrupted
Lots of missing data
Hard to analyse