Sem título

Página 1

Psychology Unit 2 - March Test Revision Notes

Aim – identifies the purpose or goal of an investigation (e.g. to investigate the effect of stereotypes on memory)

Variable – something that can change

Independent variable – directly manipulated to test the effect on the DV

Dependent variable – a measurable outcome of the effect of the IV

Extraneous variable – variables other than the IV that may influence the DV (3 types: situational, experimenter, participant)

Situational variable – environmental factors that could affect the DV

Experimenter variable – factors in the behaviour of the experimenter that could affect the DV

Participant variable – characteristics of the participant that could affect the DV

Confounding variable – any variable other than the IV that has had an effect on the DV

Operationalise – define and standardise

Hypothesis – statement or prediction of the outcome of the research

Alternative/Experimental hypothesis – predicts the effect of the IV on the DV

Directional hypothesis – predicts the direction of the relationship

Non-directional hypothesis – states that one will affect the other but doesn’t predict a direction

Null hypothesis – statement of no effect (e.g. A will have no effect on B)

Quantitative data – numerical, calculations, graphs, statistical analysis improves validity, objective – more reliable, doesn’t explain, difficult to analyse thoughts.

Qualitative data – non-numerical, descriptice, offers reasons, detailed reasons (improves validity), easier way of gathering opinions, subjective (difficult to analyse), low internal reliability due to the need for interpretation.

Primary data – gathered from source by the researcher, Milgram, Watson & Rayner, questionnaires, observations, content analyses

Secondary data – already gathered, government statistics etc.

Meta-analysis – pooling data on a particular topic, secondary data, Myles & Diener

Relative cost – PD is expensive to obtain, SD is cheaper because it already exists

Validity – PD uses careful operationalisation of variables (high validity), SD gathered for another purpose which can bring subjectivity (low validity)

Credibility – PD is more trustworthy (more valid, objective, scientifically gathered, more credible), SD less credible (collected for a different purpose)

Time Period – PD gathered at the time of the study, SD gathered some time ago so less relevant and less valid

Summary:

Strengths

Weaknesses

Primary Data

Data collected first hand for the study

Careful operationalisation of variables (high validity)

More trustworthy (collected for the purpose of the study)

Gathered at the time of the study

Expensive to obtain

Difficult to organise the study

Secondary Data

Data used by researcher, gathered by someone else

Cheaper to obtain

Can access data on groups that are difficult to study

Subjectivity (lowers validity)

Less credible (collected for a different purpose)

Less relevant data (collected previously)

Method summary:

Strength

Weakness

Lab Experiment

Lab based experiment where the IV is manipulated and all other variables are attempted to be controlled. Causal relationship identified.

Lots of control over extraneous variables

High internal validity

Artificial environment may cause behaviour to be less natural

Lowered ecological validity

Results less applicable

Field Experiment

IV manipulated but in a natural environment

Less artificial

Less likely to produce artificial behaviour

Increased ecological validity

Difficult to control extraneous variables

Quasi Experiment aka Difference studies

Not a true experiment as the researcher cannot manipulate the IV (e.g. a study on gender or Raine et al (1995))

Controlled conditions

Same strengths as lab experiments

Improved internal validity

Limits the effects of extraneous variables

Can’t be certain that the IV has affected the DV as it is not completely random

Lower in internal validity than lab experiments

Natural Experiment

Experiment in a natural environment where the IV varies naturally and the DV is observed

Allows research where it would be unethical to manipulate the IV

Difficult to control

Extraneous variables

Lowered internal validity

Cause and effect statements impossible

Participant Observation

Observation where the observer does not directly manipulate the situation and joins the group in question

Natural environment

Covert observations make behaviour more natural

Increased ecological validity

Observer bias

Bias makes results invalid

Insight may result in bias

Non-Participant Observation

Observation where the observer does not join the group – overt or covert

Researcher is not involved

More accurate observation

Improved internal validity

Observer bias

Bias makes results invalid

Content Analysis

Observing behaviour by analysing products made by others

High ecological validity

Real world observations

Observer bias

Bias makes results invalid

Questionnaire (self-report method)

A set of written open or closed questions

Large sample

Cost effective

Unassisted completion

Low response rates

Findings aren’t as valid as they could be with a larger sample

Structured Interview (self-report method)

An interview with a series of open and closed questions with no additional questions allowed

Easy to replicate

Easy to quantify

Easy to test for internal and external reliability

Not flexible

Reduced quality of data

Lowered internal validity

Semi-Structured Interview (self-report method)

An interview with general topic areas and places for the researcher to ask spontaneous questions

Qualitative data

In depth answers

Develops a sense of the person’s understanding

Interviewer bias

Bias makes results invalid

Spontaneous questions may be leading questions

Correlation

Relationship between two variables, number between -1 and +1, establishes the strength and direction of a relationship

Useful for showing strength

Identifies a need for further research

Do not show a causal relationship

Not able to identify extraneous variables

Case Study

Detailed study of a single individual, institution, or event, generally longitudinal, many data sources and collection method.

Idiographic

In depth data

Rich qualitative data

Raises internal validity

Recollection of past experiences

Retrospective data

Lowered internal validity

Correlation co-efficient – number identifying the strength and direction of the relationship

Time sampling – recording the behaviour presented after every time increment (e.g. every 30 seconds)

Time

Behaviour

0:30

Continues to work

1:00

Checks phone

1:30

Eats

2:00

Lies down

Event sampling – recording the number of times that an event occurs in a certain time period

Event

Tally

Checks phone

||||

Sleeps

||

Continues to work

||||

Eats

|||

Observations often use a behaviour checklist. For example:

Event

Checklist

Checks phone

Sleeps

Continues to work

√

Eats

Experimental designs:

Advantages

Disadvantages

Repeated measures design

Same participants take part in all conditions of the study

Eliminates individual differences

Easier to identify the effect of the IV on the DV

Higher internal validity

Order effects

Lowers the internal validity

Independent groups design

Different participants used in different conditions

No order effects

Higher validity

Lack of matching reduces internal validity as extraneous variables aren’t eliminated

Matched pairs design

Different participants used in different conditions but matched using age, gender, and ethnicity for the purpose of comparison

Individual differences eliminated

Higher internal validity

Time consuming

Counterbalancing – half do A then B, other half do B then A to improve the validity

Location of research:

Advantages

Disadvantages

Lab Environment

Controlled artificial conditions where extraneous variables are controlled

High internal validity

High control

Specialist equipment can be used

Easier to replicate conditions

Demand characteristics

Some behaviours only occur in natural settings

Low ecological validity

Field Environment

Studies taking place in natural environments (outside the lab)

Higher ecological validity

Can investigate behaviours more difficult to examine in a lab

Lower control

More likely to result in a lack of valid consent

Online Environment

Studies conducted over the internet (e.g. surveys)

Quick

Avoids observer bias

Large sample sizes

Modern

Easy

Valid consent

Open & closed questions

More unreliable

Less personal

No room clarifications

Target population – entire group that a researcher is interested in (e.g. female, lower sixth students at a college)

Sampling frame – a list of those who can be sampled (e.g. in a study about UK householders, we might choose the electoral roll as our sampling frame)

Order: identify TP, identify SP, choose sample size, choose sampling method, collect data

Advantages

Disadvantages

Random Sampling

Everyone has an equal chance of being selected (e.g. drawing from a hat)

Equal chance of being chosen

Free from bias

Unlikely to be representative

Lowered external validity

Can’t be generalised if unrepresentative

Self-Selected (Volunteer) Sampling

People come forward and offer their participation

Little effort to convince participants

Cost effective

Time effective

Unlikely to be representative

Lowered external validity Can’t be generalised if unrepresentative

Opportunity Sampling

People who are available at the time

Less time consuming

Cost effective

Unlikely to be representative

Lowered external validity

Can’t be generalised if unrepresentative

Quota Sampling

Selected by researcher according to their frequency in the population

Representative

Higher external validity

Can be generalised

Subconscious researcher bias

Stratified random sampling

Chosen randomly based on the frequency of that group in the general population

Representative

Higher external validity

Can be generalised

Time consuming

Potentially costly

Systematic Sampling

Allocating each person a number and choosing every nth number

Free from bias

Researcher has no direct influence on who takes part

Unlikely to be representative

Lowered external validity

Can’t be generalised if unrepresentative

Snowball Sampling

Choosing a small initial group and asking them to recruit other participants

Allows the study of difficult to access groups such as the homeless

Unlikely to be representative

Lowered external validity

Can’t be generalised if unrepresentative

Mean = Total of values / # of values

Median à put in order & find middle

Mode = occurring most often

Range = biggest – smallest

Standard Deviation = take mean from each value, square each result, add all totals, divide by number of values -1, square root

Measures of dispertion à range & SD

Measures of central tendency à median, mode, mean

Levels of measurement/data – types of data

Data Level

Description

Example

Nominal

Frequency headcount

Separate categories

Simplest form

_____ people helped

_____ people did not help

Ordinal

Order

Lowest to highest

Not even ranking system

1 - strongly disagree

⤋⤋⤋

5 – strongly agree

Interval

Measurements equal on a scale

Can have negative values

Temperature in ∘C

Ratio

True zero point

Equal intervals

Can have negative values

Most precise

Length in CM or

time in seconds

Graphical representation:

Internal Validity – the extent to which the IV can be said to have caused the DV

External Validity – the extent to which the results can be generalised

Ecological Validity – the extent to which the results can be generalised from the environmental conditions

Issue

Type affected

Most likely in studies where…

Resolution

Researcher Bias

Where the researcher’s beliefs influence the outcome of the study

Internal

The researcher is aware of the aims/hypothesis

Double blind study – participant and researcher unaware of aims

Social Desirability Bias

Where participants give answers that will be socially deemed as right

Internal

Participants complete surveys involving sensitive situations

Awareness of the issue

Avoid questions leading to SDB

Observational studies

Demand Characteristics

Where participants give they give the answer they expect is wanted

Internal

Participants are aware of the study

Participants are not deceived

Deceive over true nature

Don’t reveal true aim

Observer Bias

Same as researcher bias but for observations

Internal

Observer is aware of the hypothesis

Double blind design

Unrepresentative Sample

Doesn’t represent society

External

Opportunity, snowball, volunteer, systematic, or random sampling used

Other sampling methods

Small Sample Size

Sample isn’t big enough to make generalisations

External

Sample is difficult to access

Research has limited time

Larger sample size

Don’t generalise

Artificial Situation

Setting is unlike real life

Ecological

Lab experiments or unusual tasks

Avoid unusual settings

Field environments

Internal reliability – how consistently an experiment measures the behaviour in the same way throughout

External reliability – how consistently a method measures over time when repeated on different occasions

Issue

Type affected

Most likely in studies where…

Resolution

Questions on a survey do not consistently measure a behaviour

Internal

Use a questionnaire or interview method

Ensuring all items measure the same behaviour

Procedures not standardised

Internal

Use poorly carried out lab experiments

Use standardised procedures (a list of tasks that all participants must undertake)

Weather changes can influence the study

Internal

Carried out ‘on the field’

Consider conducting the study indoors or in a lab

Influence of the time of day

Internal

Not studied at same time of day

Ensure all participants are studied at a similar time

Different observers see and record different behaviour

Internal

Multiple observers used

Use the inter-rater reliability method

Get observers to agree

Difficulty with replication due to:

Ethical reasons

Obtaining samples

Poorly explained procedures

Use of case studies

External

Go against modern ethics

Use difficult to access groups

Procedures not fully reported

Use case studies

Consider an alternative design

Use snowball sampling

Conduct new study

Seek out other studies to replicate with other methods

Ethical issues:

Issue

Description

Lack of confidentiality

Personal data not protected

Participants can be found from information

Pseudonyms will resolve this

Lack of a right to withdraw

Participants aren’t clearly given the information that they can leave the study at any point with no loss

Especially important if deceived

Deception

Not told important information

Not given correct information

Cannot give fully informed consent

Often needed to prevent demand characteristics

Risk of psychological or physical harm

Not ensuring the participants leave the study in the same mental and physical state as they entered it with

Not always predictable

Risk to beliefs, values, relationships, status, or privacy

Must judge the potential risk to any of the participants’ beliefs, values, relationships, status, or privacy as a result of participating in the study

Lack of valid consent

Not fully explaining the concept and demands of the study in order for the participant to make the unbiased decision that they would like to participate

Vulnerable people (according to the BCS):

Children under the age of 16

People with learning or communication difficulties

People in care or the health service

People in custody or on probation

People engaged in illegal activities such as prostitutes

Vulnerable people should have substantial time to consider the study and form valid consent

Methods that maximise understanding should be used where possible

BCS guidelines for children:

Under 16s can’t provide valid consent

Consent must be from a legal parent/guardian

Maintain the right to withdraw

Resolutions for ethical issues:

Resolutions

Description

Debriefing

A post research interview to make the participants aware of the true intentions of the study and limit harm

Alternative types of consent

(e.g. presumptive consent)

Presumptive consent = asking a sufficiently similar group of people whether they give would consent or not

Ethical Guidelines

Professional bodies such as the BPS produce guidelines

4 principles should be considered – respect, competence (continuation of high standards), responsibility (to clients), integrity (honesty, accuracy, clarity)

Ethical committees

Within all institutions

Ensure guidelines are followed

Responsible to check research

Weigh up benefits against costs

Working with Animals

Animals (Scientific Procedures) Act (1986) governs any scientific procedure that may any distress to animals. It covers any non-human vertebrates and a single invertebrate species. It requires that all research takes place at licensed labs with licensed researchers on licensed projects.

To get a license:

Results must be important enough to justify the use of animals

Can’t be carried out with humans or computer models

Minimum amount used

Suffering is minimised through painkillers and anaesthetics

Animal Welfare Act (2006) – psychologists have a more general duty of care to protected animals. This is a duty to care for the animal properly

The Treaty of Lisbon (2009) – EU declaration that all animals are sentient and feel things

Home Office (2000) introduced the 3 Rs: Reduce, Replace and Refine

Reduce – minimise the number of animals used

Replace – seek to replace with computer models etc where possible

Refine – refine procedures so that animal suffering is minimised e.g. use anaesthetics

Responsibility to care for animals extends beyond those being studied and covers those being kept in the facility too.

Dealing with Ethical Issues:

Ethical Issue

What is it?

Resolution

What is the resolution?

Lack of Valid Consent

Not fully explaining the concept and demands of the study in order for the participant to make the unbiased decision that they would like to participate

Briefing

Telling the participants what they will be doing during the study. Reminding of a right to withdraw.

Debriefing

Post-research interview to inform participants of the true nature of the study

Presumed Consent

Asking a similar group if they would participate and using that to assume

Lack of a right to withdraw

Participants aren’t clearly given the information that they can leave the study at any point with no loss

Especially important if deceived

Informing p’s of their right to withdraw

Telling participants they can leave at any point with no loss

Debriefing

Post-research interview to inform participants of the true nature of the study

Psychological or Physical harm

Not ensuring the participants leave the study in the same mental and physical state as they entered it with

Not always predictable

Ethical Committees

Consult committees to ensure that the study is within guidelines

Ethical Guidelines

Using a professional body’s outlines to identify issues

Confidentiality Issues

Personal data not protected

Participants can be found from information

Pseudonyms will resolve this

Pseudonyms

Using false names so data cant be associated with participants

Deception

Not told important information

Not given correct information

Cannot give fully informed consent

Often needed to prevent demand characteristics

Debriefing

Post-research interview to inform participants of the true nature of the study

Ethical Committees

Consult committees to ensure that the study is within guidelines

Issues with Vulnerable groups

Allowing enough time for them to consent to the (appropriate) method

Parental Consent

Achieving consent from a responsible individual in place of the participant

Risk to beliefs, values, relationships, status, or privacy

Must judge the potential risk to any of the participants’ beliefs, values, relationships, status, or privacy as a result of participating in the study

Careful Planning

Planning in a way to avoid these risks

Pilot Study

An initial run-through of the study with a few participants

Kohlberg:

Location and Evaluation

Lab Environment

Chicago, USA

Limits Extraneous Variables

Research Method and Evaluation

A series of semi-structured interviews

In depth, quantitative data

Interviewer bias

Leading/guiding questions possible

Sample Used and Evaluation

75 boys from Chicago, USA

Androcentric

Ethnocentric

Sampling Method and Evaluation

Opportunity Sample

Lack of variation and representation

Possible similarities

Experimental Design

N/A

Variables

IV: age

DV: morality stage

Hypotheses

Alternative: The age of the participants will affect their morality

Null: Morality will not be affected by the age of the participant

Ethical Issues

Psychological harm – stress

Vulnerable Individuals

Dealing with Ethical Issues

Debriefing to remind participants of the artificiality (reduces the risk of psychological harm)

Data Types Used

Qualitative à interviews

Quantitative à grouped for the purpose of statistics

Data Sources Used

Primary

(expensive but more reliable and valid)

Validity Issues

Androcentric & Ethnocentric (external)

Controlled environment (ecological (bad) & internal (good))

Reliability Issues

Semi-structured interviews create issues in standardisation and replication of procedures

Level of Measurement (DV)

Nominal

(grouped into stages)

Milgram:

Location and Evaluation

Psychology Lab

Yale University, USA

Controlled environment (high internal validity)

Low ecological validity

Research Method and Evaluation

Controlled Non-Participant Observation

Extraneous variables controlled

High internal validity

Low ecological validity

Chance of observer bias

Sample Used and Evaluation

40 male participants, aged 20-50, varying professions

Ethnocentric and androcentric

Sampling Method and Evaluation

Volunteer Sample

Limits consent issues

Not necessarily representative

Experimental Design

N/A

Variables

IV: Verbal prods from the experimenter

DV: Level at which the teacher was willing to shock the learner

Hypotheses

Alternative: Participants will continue to ‘shock’ the confederate when the experimenter encourages continuation

Null: The verbal prods of the experimenter will have no effect on the obedience of the ‘teacher’

Ethical Issues

Deception

Psychological harm

Lack of a right to withdraw

Physical harm (seizures)

Dealing with Ethical Issues

Confirm ethical issues with surrounding psychological harm with an ethical committee and guidelines

Data Types Used

Quantitative: numbers and stats

Qualitative: filmed study

Data Sources Used

Primary Data

More reliable and valid

Expensive and time consuming

Validity Issues

Androcentric and ethnocentric (external)

Controlled environment (internal)

Controlled environment (ecological)

Reliability Issues

Difficulty replicating the study due to ethical issues in the psychological harm

Would have to consult ethical committees and guidelines

Filmed

Standardised Procedures

Level of Measurement (DV)

Nominal

­_____ men continued until the end