Research Method + Stats Flashcards
What is a pseudoscience?
a claim/belief that is presented as scientific but doesn’t use the scientific method, lacks empirical evidence and cannot be reliably tested
it is more like advice
What is psychology the science of?
the scientific study of the human mind and behaviour
how we think, feel and act
what challenges does psychology have as a science?
much of what we are interested in is unobservable
there is a question of whether we can really study human behaviour without being subjective
give a brief history of psychology as a science
- Freud - claimed to be scientific but relied heavily on introspection
- Behaviourists - only study things that are directly observable, only focussed on the S –> R
- Cognitive - makes predictions then subjects them to empirical testing, no need for direct observation
- Neuroscience - look inside the human mind and directly observe what’s happening
What is induction?
evidence gathered from multiple observations
however this cannot guarantee that it will happen again in the future
What is falsifiability?
Who came up with the concept?
the need to have the possibility to disprove a theory
if we find supporting evidence, then the theory is upheld as undefeated
Karl Popper
What is a scientific conjuncture?
a scientific question that must be falsifiable
What is Bayesianism?
The idea that beliefs come in degrees
Can express the likelihood of future events based on past knowledge
Provides a measure of a state of knowledge
What are the characteristics of a good scientist?
- Uncertain
- Sceptical
- Open-minded
- Cautious
- Ethical
What are some limitations of science?
You can’t answer questions about:
- Value
- Morality
- The supernatural
What is the hypothetico-deductive method?
- Identify behaviour of interest and generate theory
- Theory is generally derived from inductive reasoning
- Generate a hypothesis from theory - needs to be objective and falsifiable
- Subject hypothesis to empirical testing to gather evidence
- Hypothesis isn’t supported = refine or abandon
- Hypothesis is supported = uphold theory as undefeated with an estimate of confidence
What was the replication crisis?
Who was a key person here?
Methodological crisis where we couldn’t replicate the results of studies
Bem showed that people were being influenced by the prospect of being published in a journal - they were finding significant results when there weren’t any to be found (false positive)
What is a random error?
- Random error – obscures the results (normally averages out)
Longitudinal design: problems and solution
o Problem – its not possible to counterbalance order
o Solution – control group
Cross-sectional design: problems and solutions
o Problem – its not possible to randomly assign ppts
o Solution – matching
What are the types of developmental design?
- Between subjects: “cross-sectional”
- Within subjects: “longitudinal”
What is pre-test post-test control group design?
- split ppts into 2 groups and manipulate the IV in one group only
- The inclusion of a CG allows us to account for any order effects that might be present
- We can then statistically control for the difference in the treatment group accounted for by the order effects
- This is a mixed design – got within and between subjects components
What is Within subjects design without counterbalancing?
- Counterbalancing order that ppts are exposed to levels of IV is not always possible
- E.g., examining the effectiveness of mnemonic training on memory performance
- The order in which ppts are exposed to levels of IV is fixed
o Likely to be differences between time 1 and time 2 other than the variable of interest - Have to be cautious about inferring causality
What is matched pairs?
- Even better than matching the groups on the basis of potentially moderating variables
- Compare individuals from the same background but who are in different levels
- However, this frequently isn’t possible. Also, always the possibility of other variables influencing that haven’t been considered
What is matching?
- Identify potentially moderating variables and match the groups on this basis
- E.g., similar age, similar weight, similar experience etc.
What is Between subjects design without random allocation (quasi-experimental)?
- The assignment of ppts to the experimental conditions is pre-determined
- E.g., comparing pre-existing alcohol consumption groups: alcoholics vs non-alcoholics
- This poses a serious problem as there are likely to be differences between the groups other that the variable of interest
- Have to be cautious about inferring causality
Your choice of subject design depends on?
Concerns –
o Between subjects – eliminates order effects
o Within subjects – eliminates individual differences
Number of ppts available -
o Within subjects designs require fewer ppts
What is a factorial mixed design?
- One IV used between measures
- The other IV uses within measures
- E.g., night shift and then with and without alcohol, day shift and then with and without alcohol
What is a Fully repeated measures factorial design (within subjects)?
- Ppts would complete all levels in the IVs
- Ppts would complete 4 different conditions if two IVs were used
What is a Fully independent factorial design (between subjects)?
- For first IV you assign ppts to one or the other condition
- For the second IV you assign ppts to one or the other condition
- Effectively end up with 4 different groups
What is a factorial design?
- Experimental designs with 2 or more IVs
What is counterbalancing?
- Split the group of ppts in half
- Level 1 then level 2, and level 2 then level 1
- Order effects will still influence the ppts performance, but the effects of that influence will be evenly spread across each level of the IV
What is random allocation?
- Ensure each ppt is equally likely to be assigned to any IV levels
- Distributes the occurrence of potential moderating variables equally among experimental conditions
- Prevents experimenter (un)intentionally biasing their results
- Enables the use of powerful stats tests that can help determine causal relationships between variables
- Could achieve by tossing a coin, pulling straws, pulling names from a hat etc.
What is a within subject design (repeated measures)?
- Ppts exposed to both levels of the IV
- Potentially moderating characteristics are kept equal across levels
- Requires fewer ppts
- Problem of order effects – effects that occur as a result of taking a test more than once (could do better on test because they’ve done it before and know what’s coming or could do worse because they’ve done the test before and are tired of it)
- Can use counterbalancing
What is a between subjects design (independent groups)?
- Ppts are assigned to one level of the IV
- All subjects are inherently different and these differences have the ability to affect the outcome (individual differences)
- Can’t eliminate the effects of these other variables
- But we can minimise the effects by spreading their influences across the different levels of the IV
- Can use random allocation
What is the subjects design?
– the assignment of participants to experimental conditions
What is the dependent variable?
- The variable that is measured
- We compare differences in the DV under the different levels of the IV
What is an independent variable?
- The variable we manipulate
- Each have at least two levels e.g., conditions
What is the experimental method?
- A research design which allows us to make causal inferences about one or more variable of interest
- The researcher manipulates one or more variables and measures the effect on other variables
- All other variables are to be kept constant
What is ratio data?
Give an example
Ratio –
- Highest level of data
- Equal intervals and a true zero point
- E.g., height, distance, time
What is interval data?
Give an example
Interval –
- Intervals between successive values are equal
- But no “true” zero point (no absence of something)
- E.g., temperature, shoe size
What is ordinal data?
Give an example
Ordinal –
- Data can be ranked along a continuum
- Intervals between ranks are not equal
- E.g., race positions, attractiveness
What is nominal data?
Give an example
- Lowest level of measurements
- Category membership
- Numbers assigned serve as labels but do not indicate numerical membership
- E.g., gender, political party, religion
What are the scales of measurement?
- Nominal (category membership)
- Ordinal (ranked/ordered)
- Interval (equal increments but not real 0 point)
- Ratio (equal increments and real 0 point)
True or False
If you operationalize a construct it becomes a variable?
TRUE
What type of definition defines a variable?
An operational definition
What type of definition defines a construct?
A theoretical definition
What is a variable?
- anything that can assume multiple values (can vary)
- An event/condition the researchers observes/measures
- Variables must be operational – explicitly stated
What is a construct?
- A construct is the building block of a theory
- Theoretical concepts formulated to serve as causal/ descriptive explanations
- Don’t directly indicate a means by which they can be measured
Hypotheses make specific predictions and must be:
- Falsifiable
- Testable
- Precisely stated (clearly defined and unambiguous)
- Rational (needs to be consistent with known knowledge)
- Parsimonious (the explanation needs to be the simplest possible)
Theories are a collection of proposals that:
- Define
- Explain
- Organise
- Interrelate
What is a hypothesis?
– a clear but tentative explanation for an observed phenomenon
- it is testable
What is a theory?
– a broad encompassing framework.
Puts a collection of propositions/hypotheses together to explain a set of observed phenomena
What is a fact?
statements that are under no dispute.
Direct observations of phenomena that occur consistently over time
What is methodological triangulation?
The convergence of findings of methodologically varying studies which can lend acceptance to a theory pattern.
What is methodological pluralism?
The use of multiple methods
What are the quantitative research approaches?
Descriptive – describe a behaviour
Relational – predict a behaviour based on its relationship with another behaviour
Experimental – allows us to determine the cause of a behaviour
What are some criticisms of qualitative research?
- can’t apply notions of reliability or validity,
- can’t generalise to other members of the population
What is qualitative research?
- descriptive, often text-based,
- not experimental (inductive approach) but still empirical,
- focuses on underlying meaning of behaviours,
- typically small samples,
- asks more open-ended questions and so allows for unexpected answers.
What are some criticisms of quantitative research?
- humans are complex so this would be oversimplifying things,
- fails to recognise the subjective nature of research with humans,
- doesn’t look in depth at differences between individuals (focuses on averages and views population as homogenous)
What is quantitative research?
- collection of numerical data,
- subscribes to hypothetico-deductive approach,
- takes place in controlled settings,
- focuses on describing, predicting and identifying causes of behaviours,
- typically uses large samples,
- focus on structured methods
What are the goals of Psychology?
- describe behaviour
- predict behaviour
- explain behaviour
- control behaviour
What is the role of teaching in Open Science?
Teaching open science – need to teach its principles and promoting the methodological practices.
Advantages – promotes best practice, increases critical evaluation research
What are the types of replication research?
Direct replication – reproduce the elements that produced the original findings (are similar findings produced in subsequent attempts?)
Conceptual replications – change at least one aspect of the original procedure e.g., changing the sample (assess whether similar findings are produced under different conditions?)
What role does replication research have in Open Science?
Replication research – process of repeating research to verify findings
Enables confidence in results and helps build theories
What is exploratory research?
Exploratory research – focuses on generating hypotheses/research questions
What is confirmatory research?
Confirmatory research – focuses on confirming hypotheses/research question
What role does preregistration have in Open science?
Preregistration – researchers encouraged to submit plans for specific research questions and analyses they will conduct prior to data collection (transparency), reduces risk of false positives.
Promotes confirmatory research
What role does reproducible analyses have in Open Science?
- need to provide materials to enable others to generate the exact same results as those reported.
- need adequate documentation to be kept in order to do this
- Good principles: clear annotations of what documents are, store original data files separately, record all steps of data processing, use open-source software
What role does APA have in Open Science?
- APA requires researches to make data available with editors for 5 years after publication - verification, analytic reproducibility
What is the benefit of research needing public access?
- Research needs public access - allows for accumulation of knowledge, increased citation of work, more media coverage, supports meta research practice
What is Open Science?
A set of practices used to overcome limitations of previous scientific research.
Concerned with reproducibility and replicability
Research practices and data needs to be transparent and accessible
What is exploratory research?
Exploratory research – focuses on generating hypotheses/research questions
What are the sources of confounding variables?
- Selection – bias resulting from the selection/assignment of ppts to the different levels of the IV
- Results in ppts who are assigned to different levels of the IV differ systematically in some way that could influence the measurement of the DV- Particular problem for quasi-experimental designs
- History – uncontrolled events that take place between testing conditions
- Maturation – intrinsic changes in the characteristics of ppts between different test occasions
- Instrumentation – changes in the sensitivity/ reliability of measurement instruments during the course of the study
Confounds can result in us measuring:
- An effect of the IV on the DV when it is not present
- No effect of the IV on the DV when it is present
How can we eliminate confounding variables?
o Random allocation/counterbalancing spreads the influence of extraneous variables (so that they do not become confounding variables
What type of validity does confounding variables threaten?
Internal validity