Research methods

Full set of revision notes for the section A of the G544 paper. 

HideShow resource information
  • Created by: Sophie
  • Created on: 17-05-13 08:47
Preview of Research methods

First 509 words of the document:

Research methods
Types of data
Nominal data: the value is a label for a category
Ordinal data: the value can be put in order or rank. Use measures of central tendency to
summarise or display the data
Interval data: the values have set intervals between them (eg. Temperature, height).
Ratio data: same as interval data but unable to have minus values
Statistics & tests
Type 1 error: false positive (accept alternate hypo but null is true). Too lenient.
Type 2 error: false negative (accept null hypo but alternative is true). Too stringent.
5% significant level
Mean: data is normally distributed, unskewed with no outliers
Median: data is skewed or has outliers (Can't use mean)
Mode: frequency data and/or clear modal is seen
Ordinal, Interval, Ratio
Design Nominal data
Chi-squared test of
Tests of association Correlation Spearman's rank
Chi-squared test of
Tests of difference Independent measures Mann-Whitney
independent samples
Repeated measures Sign test Wilcoxon's matched pairs
Informed consent: pps should be asked if they wish to take part in the study and should be given all
relevant information about what it will involve and study aims
Deception: pps shouldn't be deceived about the aims of the study and shouldn't be deliberately mislead
about any aspect of the study
Right to Withdraw: pps should have the option to drop out at any time completely during and after the
data collection
Protecting participants from harm: pps shouldn't be harmed in any way (mentally and/or physically)
Debrief: At the end of the study, pps should be told what was happening, asked if they had any concerns
and given any explanations they require
Confidentiality: pps data and information about then should not be shared with other people who aren't
directly involved in the research
Sampling method
Opportunity sampling: pps are chosen from who happens to be around at the time
Quick and easy. Effortlessly get a large sample quickly
Unlikely to be representative of the target population (encounter people more like you)
Random sampling: pps all have an equal chance of being picked
More likely to be representative than opportunity and self-selecting
Time consuming to get right sample
Unless target population is small, it won't be easy to achieve a genuine random sample
People picked may not want to participate, need to be replaced and could lead to biased sample
Self-Selecting sampling: pps put themselves forward to take part in the study (eg. Averts)
Useful when needing a specific sample
Ppps are willing to take make and makes the sampling method more ethical
Could be expensive and more effort
People who volunteer to take part may be different in some ways (eg. More time, more extroverted,
have a particular story to tell)

Other pages in this set

Page 2

Preview of page 2

Page 3

Preview of page 3

Here's a taster:

Key details
Shows cause and effect (one thing changes, the other thing changes)
Independent variable (cause) and Dependent variable (effect)
Randomisation: used to decide the allocation of the participants into groups (eg. Coin toss)
Counterbalancing: used in repeated measures design to help overcome order effects. It's when half the
pps complete condition A then B and the other half complete condition B then A.
One-tailed: VARIABLE 1 (measured by...) will be significantly higher/lower than VARIABLE 2 (measured
by...…read more

Page 4

Preview of page 4

Here's a taster:

Reduces demand characteristics ­ may be unaware of study
Experimenter doesn't intervene directly in the research situation
Low control over IV since experimenter can't control directly
Reliability & Validity
Internal validity: establishes whether the IV was really caused by the results
External validity: effects can be generalised to different situation and to different people
Face validity: appears at face value to be measuring what it's supposed to
Concurrent validity: comparing a new test to an existing validated test
Predictive validity: can the test predict future…read more

Page 5

Preview of page 5

Here's a taster:

Behaviours won't be missed
If lots of behaviours happen at once, it may be hard to collect everything
Covert and overt
Covert: participant is unaware of observation
High ecological validity
Low demand characteristics
Ethical issues (deception)
Overt: participant is aware of observation
Avoids ethical issues
High demand characteristics
Coding scheme
A way of categorising behaviour so they can be put into categories
It clearly states when to count the behaviour (operationalized)
Tally how often the behaviours on the coding scheme appear
Inter-rater reliability
Having more…read more

Page 6

Preview of page 6

Here's a taster:

No IV or DV, only co-variables
Cannot conclude cause and effect
Hypotheses: must use correlation, not `difference'
Data could be collected from secondary source (MET office, data tables), through
questionnaires/self-reports or through observation studies
Presented in scatter graphs
Types of correlation
Positive: one variable increases, so does the other. Perfect positive correlation has co-efficient of +1.
Negative: one variable increase, the other decreases. Perfect negative correlation has co-efficient of -1.
Correlation coefficient: number between -1 and +1 and states how strong a correlation it.…read more


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all resources »