FC Y12: Part 2

List methods of collecting quantitative data?

• Experiments

• Correlational research (concordance rates)

• Structured observations

  • Behaviour categories

• Structured interviews

  • Closed questions

• Questionnaires

  • E.g: rating scales

Outline what is meant by quantitative data, and briefly evaluate it?

Numerically expressed data.

• Open to statistical testing.

• Easy to represent in charts, graphs, percentages etc…

• Tends to be more objective, less open to interpretation.

• Easy to compare

Evaluate quantitative data?

Strengths:

• Easy to analyse statistically.

  • Descriptive statistics or inferential tests.

  • Comparisons and trends are easy to identify.

• Objective

  • No researcher bias

Limitations:

• Lack of representativeness.

  • Responses gained can be narrow in their scope for explaining complex human behaviour.

  • Numerical findings can often lack meaning and context.

  • May not be a true representation of real life.

• Thus, lacks external validity.

List methods of collecting qualitative data?

• Transcripts from interviews

• Questionnaires

  • Open questions

• Extracts from diaries

• Counselling records

• Unstructured observations

Outline what is meany by qualitative data?

Language-based data

• Expressed in words.

• Written descriptions of thoughts, feelings, opinions etc…

• Can be converted to numerical data for analysis.

• Although closed questions technically give qualitative responses- it is quantitative data.

Evaluate qualitative data?

Strengths:

• Rich detail

  • So, investigator gains a meaningful insight into human thoughts, experiences and emotions.

• Increased external validity

  • Findings are more likely to represent an accurate, real world view.

Limitations:

• Subjective

  • Interpretations of this data can often rely on the opinions. and judgements of the researcher.

  • Therefore, any preconceptions a researcher has, may act as bias toward any conclusions drawn.

• Thus, researcher bias affects results.

Outline what is meant by correlation?

Correlations illustrate the nature of an association between two variables.

• This is plotted on a scattergram.

• A strong correlation is 0.7+ usually.

• You cannot infer causation from correlation.

  • Third party variables must be considered.

Evaluate the strengths of correlation?

GOOD PRELIMINARY TOOL:

• Assessing strength and direction of relationships could suggest ideas for future research.

• Provides starting point to assess patterns between variables before researchers commit.

  • Saves time, and effort.

ECONOMIC FORM OF RESEARCH:

• Quick and easy to carry out.

• No need for controlled environment or manipulated variables.

• Secondary data can be used.

  • So, even less time consuming.

Evaluate the limitations of correlation?

ONLY TELLS US THAT VARIABLES RELATE AND NOT WHY:

• Cannot demonstrate cause and effect, so we cannot be sure which variable is confounding.

  • e.g: caffeine or anxiety.

• The direction of correlation could be reversed.

MISUSE OF RESULTS:

• The issues above can lead to misused and misinterpreted results.

• Relationships often also are stated as facts, when they are not.

• This can lead to prejudice and stereotypes.

ISSUE OF THIRD VARIABLE

Define reliability?

Getting the same results by repeating the study.

How could reliability be improved in self reporting questionnaires?

• Avoiding complex or ambiguous language.

• Replacing open questions with closed questions.

• Having preset questions.

How could reliability be improved in interviews?

• Having the same interviewee for everyone.

• Training the interviewee.

• Structured interview style.

• No leading or double barrelled questions.

List the two ways of assessing reliability?

• Test re-test

• Inter rater

Describe the process of inter-observer reliability?

1) All observers decide on behaviour categories together, then familiarise themselves with them.

2) They observe the same things at the same time but separately.

3) Compare results.

4) Correlational analysis. 0.8+ is reliable.

Describe test-retest reliability?

Administer the same test or questionnaire to the same person on a different occasion.

• A reliable test should give the same results.

Briefly evaluate closed questions?

Pros:

• Produces both quantitative and qualitative data.

• Easy to analyse

• More reliable

Cons:

• Fixed number of responses available

  • (e.g: good, bad etc)

• Lacks detail

Briefly evaluate open questions?

Pros:

• Qualitative data- so very rich in detail

• Usually more valid

Cons:

• No fixed range

• Can be answered vaguely and in any way

• Subjective, so answers are hard to categorise

Define questionnaires

A pre-set list of questions (items).

• Often used in experiments to assess the dependant variable.

Evaluate questionnaires?

Strengths:

• Cost effective

• Little effort

• Data easy to analyse

• Data is perfect for statical analysis

  • Particularly if closed questions used.

Weaknesses:

• Response bias

• Social desirability bias

• Acquiescence bias

List and describe 3 ways of designing a questionnaire?

• LIKERT SCALE: (e.g)

  • 1- strongly agree 2- agree 3-neutral 4-disagree 5-strongly disagree

• RATING SCALE: (e.g)

  • High educational value- 1 2 3 4 5 - Low educational value

• FIXED CHOICE OPTION: (e.g)

  • Why do you watch films?

    • education

    • fun

    • interest

What is response bias?

When people cannot be bothered to complete a questionnaire fully, so randomly tick the same box etc.

What is social desirability bias?

When people answer in an unnatural way to please others or fit in with what is considered ‘normal’.

What is acquiescence bias?

When people are more likely to say yes than no.

Define interviews, and the 3 types?

A live encounter, usually face to face.

STRUCTURED:

• Predetermined questions.

• No follow ups

UNSTRUCTURED:

• Conversational, free flowing

• No set list of questions

SEMI-STRUCTURED:

• Preset list of questions but follow ups are allowed.

List and describe three things you need to avoid to write good questions for interviews?

Avoid jargon:

• The overuse of jargon is unnecessarily complex and confuses people.

Avoid leading questions and emotive language:

• These push people into certain opinions and forces them to agree.

  • This reveals the interviewee’s opinion.

Avoid double barrelled questions and double negatives:

• Double negatives can be confusing.

• Double barrelled questions push you into a point of view, making it harder to disagree.

Describe 5 things that should be taken into account when designing interviews?

• Interview Schedule:

  • Pre set list of questions if structured.

• Standardised:

  • (Posture tone of voice etc…)

  • Reduces confounding effect of interviewer bias.

• Establishing rapport with neutral questions:

  • Helps participant to open up.

• Single or group interviews:

  • Quiet room, away from others.

• Reminder of ethical issues:

  • Confidentiality, anonymity.

How can social desirability bias be minimised in interviews?

By establishing a rapport

Evaluate structured interviews?

Strengths:

• Straightforward to replicate.

• Standardised

Weaknesses:

• Less flexible

• Cannot deviate from the topic

Evaluate unstructured interviews?

Strengths:

• Lots of flexibility

Weaknesses:

• Analysing data is harder

• Not standardised

What is a weakness of both structured and unstructured interviews?

Social desirability bias

List the 6 types of observation?

• Naturalistic

• Controlled

• Covert

• Overt

• Participant

• Non-participant

Why must researchers not conduct observations alone if possible?

It can be biased and subjective.

Describe naturalistic observations?

Takes place in a setting where the behaviour would normally occur.

• All aspects are free to vary.

• Often used to observe manager-employee interactions.

Evaluate naturalistic observations?

Strengths:

• High external validity

• Generalisable to real life

Weaknesses:

• Lack of control makes replication difficult

• Patterns of behaviour can be hard to judge due to EV’s.

Describe controlled observations?

Variables are manipulated to observe effects.

• EVs are controlled.

• Strange situation study is an example.

  • (two way mirror)

Evaluate controlled observations?

Strengths:

• Well controlled

• Replication is easier

Weaknesses:

• Lacks external validity

• Not generalisable to real life

Describe participant observations?

The researcher themselves takes part in the study.

• Provides first hand account of environment.

Evaluate participant observations?

Strengths:

• Provides more insight

• Increases external validity

Weaknesses:

• Researcher might begin to identify too strongly with the group or risk going ‘native’.

• Thus, objectivity is lost.

• It can also be inappropriate for a researcher to take part.

  • For example, an adult researching children’s behaviour in a playpark.

Describe non-participant observations?

Researchers remain separate when conducting research.

• More objective

Evaluate non-participant observations?

Strengths:

• Objective

Weaknesses:

• Lose valuable insight

• Thus, decreased external validity

Describe covert observations?

Participants are unaware they are part of the study and being observed.

• Behaviour must be in a public place and be happening anyway to be legal.

• Behaviour may be recorded without gaining consent.

Evaluate covert observations?

Strengths:

• Removes demand characteristics

• Increased internal validity

Weaknesses:

• Ethical issues

• Lack of proper consent

Describe overt observations?

Participants are aware that they are part of the study.

• They have given fully informed consent.

Evaluate overt observations?

Strengths:

• Ethically acceptable

Weaknesses:

• Leads to demand characteristics

• Reduced internal validity

Briefly describe and evaluate unstructured sampling techniques?

Continuous readings used.

So, all instances of target behaviour are recorded.

Pros:

• Lots of rich detail

Cons:

• Risk of observer bias

  • They might only observe things that ‘catch their eye’.

• Can be an overwhelming amount of detail, making analysis difficult.

List two types of structured sampling?

• Event sampling

• Time sampling

Briefly describe and evaluate structured sampling techniques?

Uses systematic methods.

Pros:

• Quantitative data

• Easy to analyse and compare

Cons:

• Not enough detail

Describe and briefly evaluate time sampling?

Involves recording behaviours within a pre-established time frame (e.g: every 2 minutes)

Pros:

• Effective for reducing the amount of observations needed.

Cons:

• Could be unrepresentative.

Describe and briefly evaluate event sampling?

Counting the number of times a specific behaviour occurs in a target individual or group.

Pros:

• Useful when behaviours happen infrequently and could be missed.

Cons:

• Complex behaviours may be missed.

Describe what is meant by behaviour categories in a structured observation?

This is when the target behaviour is broken down into specific categories.

• This leaves no room for inference.

• The observer must include all the ways behaviour can occur but without overlap.

  • affection= hugging, kissing, holding hands

Evaluate behaviour categories?

Pros:

• Makes data collection easier

• More objective

Cons:

• Must be no overlap

  • e.g: smiling, grinning

  • as it can lead to ambiguity!

Describe unstructured observation as a method of design?

The researcher quantifies observations by writing down everything they see.

• It is often small scale with fewer participants!

Evaluate unstructured observations?

Pros:

• Richer

• More detailed

Cons:

• Produces qualitative data which is harder to analyse.

• Observer bias, as they only record what ‘catches their eye’ due to absence of strict behaviour categories.

Describe structured observation as a method of design?

Researchers quantify their observations using a predetermined list of behaviours and sampling methods.

Evaluate structured observation as a method of design?

Pros:

• No observer bias due to strict behavioural categories.

• Often produces quantitative, numerical data.

• Makes analysis of data easier.

• Recording of data is easier.

Cons:

• Less detail produced.

• Strict behavioural categories.

Describe inter rater reliability?

Researchers should work in teams of 2 and complete the study separately.

Their results should be correlated at the end to assess the reliability.

Define a non experimental method?

One that cannot determine cause and effect.