# Research Methods

?

## Volunteer Sampling

- When the participants select themselves to take part in the study

* Easy and time efficient
* All participants will have consented
* May produce a large sample

Sample may be biased - May be the same type of people who volunteer e.g students
Sample may not be valid
Sample may not be representative
Sample may show demand characteristics

1 of 36

## Opportunity Sampling

- Sample collected by asking individuals who are available at the time and fit the criteria you’re looking for.

* Time efficient
* Cost effective
Sample likely to not be representative May try to be seen as socially acceptable/demand characteristicsResearcher bis when selecting

2 of 36

## Random Sampling

- Each member is chosen by chance and everyone has an equal chance of being selected.

* Avoids bias and researcher has no control * Law of probability says the sample is likely to be representative

Time consuming
Chance the sample may not be representative

3 of 36

## Experiment

A SITUATION LOOKING AT CAUSE AND EFFECT AND INVOLVES MANIPULATING A VARIABLE.

4 of 36

## Hypothesis

- A statement which can be tested to see whether or not it is true.

Research: general prediction ( not enough info to base an investigation on)

Alternate: Enough datail for the experiment to be carried out as components are operationalised. Meaning the variables are measurable

ONE TAILED: "There will be a significant increase..." The hypothesis has a clear direction.

TWO TAILED: "There will be a significant difference..." The hypothesis has no clear direction.

NULL: "There will be no significant difference..." The hypothesis predicts no difference whatsoever.

5 of 36

## Variables

Independent Variable: The one you manipulate
Dependent Variable: The one you measure

Extraneaous Variable: Any factor excluding the independent variable which could affect the results (dependent variable).

Confounding Variable: A variable which cannot be controlled. (May do unethical to do so, or mya just be impossible).

6 of 36

## Bias

Demand characteristics: Subjects try to make sense of the study they are participating in and adjust their behaviour accordingly.

Social desrability: Subject gives a response which fits in best with the general opinion or view of their peers.

Experimentor bias: The experimentor's expectations or preferences influence the outcome of a study.

Observer Bias: The presence of an observer may change the behaviour of those being observed.

7 of 36

## Validity

Whether or not the study measures what it intended to measure.

Internal Validity: Whether the effects observed in the study are due to the manipulation of the independent variable and not another factor.

External Validity: Refers to the extent to which the results os a study can be generalised to other settings.

Face Validity: Whether the measuring tool appears to be doing what it should.

Ecological Validity: Whether the behaviour measured is representative of behaviour that naturally occurs.

Population validity: How well the sample can be used to generalise to the population as a whole.

8 of 36

## Reliability

The consistency of findings: if the test was repeated using same method would the same results be achieved?

Internal reliability: How consistently a method measures within itself (using one test)

External reliability: The consistency of results over time which repeated

Inter-rater reliabilty: The consistency of different raters working on the same study within their findings

Test re-test method: Participants take same test of different occasions. a high correlation between test scores shows high reliability

9 of 36

10 of 36

## Field Experiment

- An experiment carried out in the natural environment

* Can generalise findings
* High in ecological validity
* IV can still be manipulated

Time consuming
Confounding variables make harder to establish cause and effect

11 of 36

## Quasi Experiment

- The experimentor does not manipulate the IV, it is naturally ocuring. Can be tested in the lab or in the field.

* High in ecological validity
* More ethical as subjects are not manipulated

Confounding environmental variables (cannot confidently determine cause and effect)
Have to wait for subjects which characteristics of IV to become available

12 of 36

13 of 36

## Independent Measures Design

Different participants used in each condition

* No order effects
* Demand characteristics reduced as subjects only used in one condition

Individual differences may influence results
More subjects needed

14 of 36

## Matched Pairs Design

Uses different paricipants in each condition however they are similar or matched on characteristics, e.g twins are useful.

* Extraneous variables are well controlled
* Individual difeerences reduced

Time consuming
Expensive

15 of 36

## Controls

Standardisation: Method and all instructions made identical for each participant.

Counterbalancing: The researcher changes the order of task for each participant or conditions to try to control or lessen the impact of order effects.

Randomisation: The order of task is randomised in an attempt to control or lessen the impact of order effects.

16 of 36

## Observation

To simply observe and record behaviour (no IV is manipulated)

17 of 36

## Covert Observation

The subjects are unaware they are being observed

* No demand characteristics as unaware of study
* High in ecological validty

Unethical
Difficult to observe without being seen

18 of 36

## Overt Observation

Subjects are aware they are being observed

* Ethical as they have given informed consent
* Easier to conduct

Demand characteristics as subjects may alter behaviour

19 of 36

## Structured Observation

Researcher divises a checklist of the predetermined behaviours they are looking for. Provides qualitative data as the recurrence of behaviours displayed is recorded using a tally.

* Easy to analyse and draw conclusions
* Reliable

Researchers may miss interesting behaviours as not on checklist
Open to researcher bias

20 of 36

## Unstructured Observation

Observer records all behaviours producing qualitative data.

* Provides the researcher with detail
* Less likely to miss important behaviours

Difficult to concerntrate for long periods of time and therefore may not witness all behaviours displayed

21 of 36

## Event Sampling

Observers have predetermined set checklist of behaviours and observe a whole event, tallying behaviours when seen.

* Quantitative data
* Easy to analyse and draw conclusions from

Difficult to concerntrate for long periods of time
May miss important behaviours if not on checklist

22 of 36

## Time Sampling

Observers watch for behaviours for set periods of time in between set intervals of time.

* More focused (Short periods of time)

Observations may not be representative
May miss important behaviours within intervals

23 of 36

## Participant Observation

Participant takes part in the behaviour they are observing

* High in ecological validity
* Reduced demand characteristics

Difficult to simultaneously record behaviour
Observer may influence behaviour

24 of 36

## Non-participant Observation

Participant does not take part in the behaviour which they are observing

* Easier to record information
* Can more easily observe behaviour

Higher demand characteristics if subjects are aware they are being observed
Researcher bias

25 of 36

## Self Report

Used to gather peoples’ opinion and ideas on a topic and to gain an insiders perspective. Administered: Face to face, by post, en masse to a group in public setting, via phone or internet

26 of 36

## Questionnaire

Asking a large sample for information on a topic at a particular moment in time

* Large sample (representative)
* Large amounts of data collected
* Efficient as researcher does not have to be there whilst completed

Social desirability bias
If untruthful responses given reduces validity
Cannot be sure who completed it

27 of 36

## Open/Closed Questions

OPEN: Allows subject to freely express view/opinion in their own words

* Qualitative data
* High valitdity
Qualitative data - difficult to analyse
Low reliability

CLOSED: Respondents must choose an option for their response

* Quantitative data
* Standardised - High reliability
Low validity if chosen response not an option
Lacks detail

28 of 36

## Structured Interview

Interviewer sticks to a strict list of questions and uses a standardised procedure

* High reliability
* Easy to compare and analyse response
* High inter-rater reliability

Lacks validity as subject not free to expand on responses
Lacks detail

29 of 36

## Unstructured Interview

Researcher has freedom to vary questions and go into more detail with responses

* High validity
* Produces more detailed information

Time consuming
Much of the information may not be relevant

30 of 36

## Likert Scale

List of attitudes about a statement and asked to indicate using a scale how strongly they agree or disagree.

* Quantitative data
* Extent of opinion measured

Social desirability bias
May misinterpret scale

31 of 36

## Correlation

Shows a relationship between two co-variables

Positive correlation: As the values of one co-variable increase, the values of the other co-variable increase also (+1 coeffiecient)

Negative correlation:
As the values of one co-variable increase, the values of the other co-variable decrease. (-1 coefficient)

No correlation: There is no relationship between the two co-variables (0)

32 of 36

33 of 36

## Descriptive Statistics

MEAN: Adding all values together and dividing by the number of values

* Most representative measure
May be influenced by extreme values

MEDIAN: The central number when values in chronological order

* uneffected by extreme values
Not all values represented

MODE: The most common value

* Can be used with any data type
Not useful with small data sets

RANGE: Difference between biggest and smallest value

*Easy to calculate
Doesn't indicate how widely or tightly spread a group of values are

34 of 36

## Standard Deviation

Calculates difference of a score from it's group mean

* More precise than range as all data accounted for
* Allows researcher to know how much scores vary amongst themselves

Difficult to calculate

35 of 36

36 of 36