Cheatography
https://cheatography.com
INTRO TO PL2131
Learn how to conduct psychological research |
- Turning a question into research
- Designing an experiment
- Collecting and analysing data
- Presenting findings
SCIENTIFIC RESEARCH – the scientific approach
1) Intuition: process of coming to direct knowledge or certainty without reasoning or inferring; forming hypotheses |
2) Authority: acceptance of facts stated by authorities; used in designing stage; expert whose facts are subject to testing using the scientific process |
3) Rationalism: uses reasoning to arrive at knowledge, assumes that valid knowledge is acquired if correct reasoning process is used; identify the outcomes that indicate the truth/falsity of the hypotheses |
4) Empiricism: acquire knowledge through experiences; cognition and perception; empirical observations to be conducted under controlled conditions |
- The goal of science: to understand the world we live in
- To acquire knowledge
ASSUMPTIONS UNDERLYING SCIENTIFIC RESEARCH
2) Reality in nature: our experiences are real; forms basis for further research; scientists assume that there is an underlying reality that they are trying to uncover |
3) Discoverability: it is possible to discover the regularities and reality; must assume that we can discover laws that make experiences real |
1) Uniformity/regularity in nature |
a. Determinism: the belief that there are causes or determinants of mental processes and behaviour (making sense of the world)
b. Probabilistic cause: causes that usually produce outcomes, the interim and what we get instead when we are seeking to attain the end goal that is determinism
PSYCHOLOGICAL RESEARCH
Conceptualisation |
adopting a scientific approach; definition of terms |
Operationalisation |
construct vs measure; working definition of the construct - specification |
Hypothesis |
forming a testable hypothesis; science is falsifiable; embracing the null; can never be proven to be correct |
Research study |
experimental vs non-experimental |
Data collection |
how do we treat subjects? measurement modes used |
Data analysis |
samples and sample sizes; comparing group scores |
Presentation |
presenting research findings |
MEASUREMENT MODES
Nominal |
categories, non-quantitative, uses symbols to classify variable values |
Ordinal |
rank-order scale of measurement; cannot assume equidistance |
Interval |
equal intervals, no absolute zero point (arbitrary) |
Ratio |
absolute zero point, rank-ordering, equal intervals |
GOOD MEASUREMENTS
- Reliability: consistency of scores of your measurement instrument
- Validity: extent to which your measurement procedure is measuring what you think it is measuring; whether you have used and interpreted the scores correctly
EXPERIMENTAL RESEARCH
Quant. exp. research designs |
Conducting experiments to establish causations by manipulating IVs and observing changes on DVs |
Required conditions for claiming causation:
- Association: 2 variables are empirically correlated
- Temporality: cause comes before effect
- Elimination of plausible alternative explanations: effect cannot be explained by a 3rd variable
INDEPENDENT VARIABLES
Levels of the IV and manipulation strength |
>2 levels of the IV to conclude causality |
Strength: levels of the IV must be distinct and different from each other |
# of IVs? |
>1 IV!! |
Having only one -> misleading |
In experimental designs:
- Event manipulation: random assign. into conditions, roughly equal profiles
- Instructional manipulation
- Individual difference manipulation: varying IV by selecting participants that differ in the amt or type of a measured internal state (cannot conclude causality; inherent characteristics)
DEPENDENT VARIABLES
In experimental designs |
They can be continuous or categorical in nature |
Number of DVs |
There can be alternatives! -> accuracy/response time |
EXTRANEOUS VARIABLES
- Third variables besides the IV and DV
- Cloud interpretations of the IV-DV rship if uncontrolled
- Blinding to remove bias (systematic ways to account for them)
EV vs CV
EV |
CV |
- Might compete with the IV in explaining the outcome |
- An EV that may eliminate the ability to claim that the IV causes changes in the DV |
- Affects absolute outcome but not experimental outcome |
- Creeps in systematically and affects one level of the IV but not the other |
DESIGNS
Between (goes through 1 level of the IV) |
Within (goes through all levels of the IV) |
Shorter time to obtain results |
Elimination of CVs |
Random assign. could cause unequal groups of unequal abilities (confounding) |
Mental fatigue, floor effects |
|
|
EXPERIMENTAL CONTROL
Between |
Within: counter-balancing to counter sequencing effects (order effects and carryover effects) |
- matching: alt. method to/can be combined with randomisation |
- randomised: possibility that there is a sequence that has a higher frequency of a certain variable |
- randomisation |
- intrasubject: does not solve order effects |
| |
- complete: N!, N = # of levels of IV; may not have enough participants |
| |
- incomplete: multiple sequences, control order effects, N sequences, only works for even #; odd # – create a mirror! |
Matching:
o Equating participants
Precision-control: each participant matched with another on selected variables (equal identical attributes);
Freq. distribution: match groups by equating overall distribution of selected variable – random assign til 2 groups comparable
o Hold variables constant: slicing
o Build the EV into research design
Incomplete:
Each TC appear equal no. of times in each position
Each TC precede and follow every other TC equal no. of times
NON-EXPERIMENTAL RESEARCH
Experimental |
Non-experimental |
manipulated the IV (variability) |
did not manipulate the IV (variability due to individual differences |
can infer causality |
can only infer correlation |
control over EVs |
construct and use good test items |
SURVEY RESEARCH METHODS
1. Match the research objectives. |
2. Appropriate for the respondents to be surveyed. |
3. Short, simple questions. |
4. Avoid loaded or leading questions |
5. Avoid double-barrelled questions |
6. Avoid double negatives |
7. Determine whether closed-ended, or open-ended, or mixed format questions are needed |
8. Construct mutually exclusive and exhaustive response categories for closed-ended questions |
9. Consider the different types of closed-ended response categories (measurement modes) – would an interval scale or ordinal scale be more useful? |
10. Use multiple items to measure complex or abstract constructs |
11. Make sure questionnaire is easy to use; - Limit contingency questions (redirection) - Control response bias (social desirability) - Control response bias (response set) – insert contrasting items |
12. Pilot-test – think-aloud technique |
Need to ensure the validity of questionnaire (i.e., the test items measure what we had initially set out to measure)
Construct is too broad for comfort: need to operationalize
Specific operationalization of the idea that we want to pursue and not something else
DESCRIBING SCORES
Mean |
Variability |
| |
- Wanting to know how the scores spread around the mean |
- Presence of outliers can be misleading |
Standard deviation: describing the spread of a group of scores; average amount that scores differ from the mean |
| |
Variance |
Central tendency:
- Make sense of a group of scores
- Know how our data look like centrally
INFERENTIAL STATISTICS
1. Converting raw scores to Z-scores |
2. Converting Z-scores to raw scores |
- Number of SDs a score is above or below the mean |
X=(Z)(SD)+M |
Z=(X-M)/SD |
Distribution of Z-scores: M=1,SD=1 |
Z-scores
- To describe a score in terms of where it fits into the overall group of scores, create a Z-score
- Number of SDs a score is above or below the mean
- Analogous to a translation; standardisation
!! We describe a group of data scores using a representative value (mean + SD)
Obtain a Z-score to infer how a score is ‘performing’ in comparison to others.
EFFECTS
Ceiling effect |
when an IV no longer has an effect on the DV |
Floor effect |
when a data-gathering instrument has a lower limit to the data values it can reliably specify |
|
Created By
Metadata
Favourited By
Comments
No comments yet. Add yours below!
Add a Comment
Related Cheat Sheets