Psychology Unit 1: Research Methods

All of Research Methods for AQA Psychology.

HideShow resource information
  • Created by: emmaaaaa
  • Created on: 05-05-14 14:40

Research Methods Definitions

  • Qualitative- data that is not numbers e.g. words, descriptions etc.
  • Quantitative- data that is numbers.
  • Gender Bias- having more of one gender than another in a sample, so cannt generalise.
  • Sampling Bias- sample not representative of population.
  • Cultural Bias- focus on one culture e.g western culture, not generalise to all population.
  • Incentive- change results if e.g. getting paid, more likely to conform.
  • Sample size- larger the sample the better.
  • Demand characteristics- participants aware being studied so adjust behaviour.
  • Independent variable- controllable variable that affects the dependent variable.
  • Dependent variable- the variable being measured, affected by independent variable.
  • Extraneous variables- other variables that could cause changes in dependent variable.
  • Reliable- results consistent, if repead experiment get same results.
  • Validity- did test measure what intended to: internal- did method work, external: generalise?
  • Ecological validity- how well the study relates to/reflects real life.
  • Ethical issues- are participants put in harms way, do ends justify means etc.
  • Debriefing- tell participants about the study after it has been conducted.
1 of 40

Lab Experiments

  • WHAT IS IT
  • Experiment in artificial environment, highest level of control, allow sophisticated equipment use.
  • carried out in Lab, often memory/cognitive studies, but wide range uses.
  • ADVANTAGES
  • Can be replicated if carried out carefully.
  • Sophisticated equipment can be used easily
  • Easier to control variables.
  • DISADVANTAGES
  • loss of valitidy
  • lack ecological validity
  • Demand Characteristics
  • ETHICAL ISSUES
  • consent
  • Deception- some studies need deception to get results, must tell true purpose after.
  • Use of Animals- cannot consent, not subject unnecessary suffering.
2 of 40

Field Experiment

  • WHAT IS IT
  • In natural setting, intend prove realism of experiment.
  • Most often used study non-human behaviours, still manipulate IV to change DV.
  • ADVANTAGES
  • Ecological validity
  • Reduction of demand characteristics
  • DISADVANTAGES
  • difficult to have full control of external factors
  • Not able to generalise to other real life situations
  • Usually time consuming
  • difficult use technical equipment restricting what can be measured.
  • ETHICAL ISSUES
  • consent
  • confidentiality
  • Animals
3 of 40

Natural Experiment

  • WHAT IS IT
  • type of quasi experiment, allocation of participants to experimental conditions is not in control of researcher, but reflect naturally occuring differences in independent variable.
  • ADVANTAGES
  • Reduction of demand characteristics
  • Lack direct intervention
  • DISADVANTAGES
  • loss of control
  • chances of desired behaviour occurring are low.
  • ETHICAL ISSUES
  • consent
  • protection from harm
  • confidentiality
4 of 40

Correlational Analysis

  • WHAT IS IT
  • Correlation relationship of 2 variables, strength range from +1 to -1 (correlation coefficient)
  • investigate extent relationship between 2 variables, used early stages research on topic to identify variables, use analyse results other experiments.
  • ADVANTAGES
  • precise quantative measure of strenght relationship
  • enables measurement many variables and relationships
  • useful try understand complex relationships, helpful exploratory research.
  • DISADVANTAGES
  • impossible establish cause and effect- CORRELATION DOES NOT EQUAL CAUSATION
  • non linear relationships cannot be measured by commonly used correlational techniques.
  • ETHICAL ISSUES
  • consent
  • right to withdraw
  • keep informed of use of research if aims change.
5 of 40

Case Studies

  • WHAT IS IT
  • Research involve in-depth study over time of case, usually individual or small group.
  • undertaken real life context, data mainly inverviews/observation, qualitative
  • some measurement data produced, powerful account overall picture
  • Several case studies referred to as multiple case study.
  • ADVANTAGES
  • rich interesting data- high degree realism, invaluable insight provided.
  • challenge existing theories- finding single study contradict theory, modify to accomodate.
  • Makes use of rare cases e.g. feral children.
  • DISADVANTAGES
  • low reliability
  • subjectivity
  • large amounts of data- difficult to process
  • Distortions- errors in recall in historical events.
  • ETHICAL ISSUES
  • privacy/confidentiality, psychological wellbeing
  • Informed consent/debriefing
6 of 40

Observational Studies

  • WHAT IS IT
  • Naturalistic observation- obsever natural setting, not change behaviour.
  • no deliberate manipulation, method used any other type inappropriate.
  • Often carried out natural setting, sometimes in lab e.g. rats
  • researcher participate or not, or be disclosed/undisclosed. (aware/unaware observed)
  • ADVANTAGES
  • value as preliminary research tool, lead further research avoid unnecessary experiments
  • careful use identify appropriate hypothesis.
  • Validity, reduction of demand characteristics
  • DISADVANTAGES
  • poor control of variables, replication difficult
  • prescence of observer change behaviour
  • costs can be high
  • codin issues
  • ETHICAL ISSUES
  • privacy, confidentiality
  • consent, data recorded used appropriately
7 of 40

Self Report Techniques- Questionnaires

  • WHAT IS IT
  • info gather specific interests/areas, carried out on samples.
  • Closed- answers grouped e.g. tick box, easy quantify to analyse, lack realism, forced choices.
  • Open ended- not restricted, answers detailed, harder to analyse cost more
  • Leading questions- wording implies participants respond certain way (should be avoided)
  • ADVANTAGES
  • very simple, quick and cheap
  • researcher has less influence
  • DISADVANTAGES
  • wrong interpretation of questions
  • leading questions influence outcome
  • answers may not be truthful- desirable answers.
  • ETHICAL ISSUES
  • privacy
  • consent needed
  • protect from psychological harm, provide support
8 of 40

Self Report Techniques- Interviews

  • WHAT IS IT
  • face to face, can form basis of case study, way forming survey
  • Structured- quantative data set questions, less likely deviate from topic, generalise easier.
  • Unstructured- start broad question, improvise, more valid, like actual conversation, qualitative.
  • Semi-structured- prepared but use additional questions, usually most successful.
  • ADVANTAGES
  • flexible, assess behaviour of people as well as answers
  • different structures suit different research
  • DISADVANTAGES
  • data misenterpreted, time and effort needed.
  • hard to interpret data
  • interviewers need training
  • ETHICAL ISSUES
  • privacy/confidentiality and wellbeing
  • right to withdraw
  • Debrief
9 of 40

Sampling

  • Sampling- participants used from the target population to collect data that could be generalised to the remainder of the target population if it is truly representative.
  • There are many ways in which a sample is chosen which include:
  • Random Sampling
  • Opportunity Sampling
  • Volunteer Sampling
10 of 40

Random Sampling

  • WHAT IS IT
  • every person/item in target population has equal chance of selection.
  • Necessary have a list of every person/item in target population to generate it.
  • Must happen in unbiased way.
  • ADVANTAGES
  • equal chance of being chosen.
  • Laws probability predict chance selecting biased sample limited.
  • DISADVANTAGES
  • not guarentee totally representative of population concerned.
  • any 2 random samples drawn same target population may not share identical characteristics.
11 of 40

Opportunity Sampling

  • WHAT IS IT
  • widely used non-random method of sampling as it is convenient.
  • select anyone available to take part in study from given population.
  • ADVANTAGES
  • convenient, quick and easy.
  • DISADVANTAGES
  • findings cannot be generalised
  • unlikely generate sample representative of wider target population
12 of 40

Volunteer Sampling

  • WHAT IS IT
  • self selected sampling, non-random sampling technique.
  • participants select themselves to take part in a research study- often reply to an advert.
  • Widely used in university research.
  • ADVANTAGES
  • easy to do, little effort
  • DISADVANTAGES
  • majority of given target population unlikely to respond to request, those who do may not be typical of target population.
  • unlikely be representative sample of target population-biased
  • Findings cannot be generalised
13 of 40

Code of Ethics and Conduct

  • Regulated by The British Psychological Society (BPS)
  • Psychologists have to defend their proposed research to an ethics comittee before allowed to proceed.
  • Enforced- guidelines differ across countries, no universal set of guidelines.
  • Effectiveness of code depends on ethics comittee enforcing it (must also include non-psychologists)
  • Most be willing punich those who break rules, BPS imposes penalties on their members, however no obligation for psychologists belong to BPS, punishments usually financial.
14 of 40

Informed Consent

  • WHAT IT IS- all participants helped to understand aspects of research likely to influence willingness to participate, nature and objectives of investigation, give consent take part.
  • HOW TO SOLVE IT
  • Children- parent or legal guardian give consent.
  • Presumptive- sample outline study to them, would they have argued with deception, presume if sample accept others will.
  • Prior general- tell using deception, but not tell them what type of deception.
15 of 40

Confidentiality and Anonymity

  • WHAT IT IS- data must be stored securely (Data Protection Act 1998), participants have right for all information gained from research to remain confidential/anonymous.
  • HOW TO SOLVE IT
  • if confidentiality cannot be guarenteed, participants must be told prior to consent given.
  • Can change names to numbers or fake names etc. to hide identity.
16 of 40

Right to Withdraw

  • WHAT IT IS- made clear that participants can withdraw at any time, without having to give a reason and any data related to them can be removed and destroyed.
  • HOW TO SOLVE IT
  • consent- remind personal questions have confidentiality/anonymity.
  • remind participants don't need to answer personal questions in specific aspects of study.
17 of 40

Deception

  • WHAT IT IS- full disclosure to participants of what research they are taking part in involves where possible, in some cases deception is needed to get results e.g. Milgrams experiment into obedience.
  • HOW TO SOLVE IT
  • clear document specifying each stage of experiment for each participant.
  • If uneasy after debriefing remind right to withdraw, must disclose everything in debriefing.
18 of 40

Observation Research

  • WHAT IT IS- respect privacy, psychological wellbeing of individuals studied
  • HOW TO SOLVE IT
  • avoid possibility of invading privacy.
  • gain informed consent
  • only acceptable public places when observed by strangers anyway
19 of 40

Risk of Harm

  • WHAT IT IS- protect participants against physical/mental harm in research, risk should be no greater than real life, consider study from participants perspective.
  • HOW TO SOLVE IT
  • consider study from ethical point of view and participants perspective.
20 of 40

Understanding Implications

  • WHAT IT IS- make sure participants understand the significance of the research that they are taking part in.
  • HOW TO SOLVE IT
  • learn as much as possible about the area researching
  • Consult an expert in the field you are studying.
21 of 40

Protection from Stress

  • WHAT IT IS- protect participants from undue stress, remind they do not have to answer some questions if they do not wish to.
  • HOW TO SOLVE IT
  • make aware that participants do not have to do anything that they do not want to
  • not investigate personal/private matters.
22 of 40

Inducements

  • WHAT IT IS- financial incentives not used encourage take part in research, if risk of harm is beyond normal life risk.
  • HOW TO SOLVE IT
  • do not use if any danger greater than real life.
  • e.g. want to torture someone, they say not but offer £200 and they change their minds.
23 of 40

Proffessional Advice

  • WHAT IT IS- responsible to inform participants if they have psychological problems brought to light by research, and give them advice on what to do.
  • HOW TO SOLVE IT
  • offer it to those who need it.
  • alert them to pscyhological problems may have been caused by study.
24 of 40

Non-human Animals

  • WHAT IT IS- highest standards of animal welfare be observed, animals not subject more main than necessary.
  • HOW TO SOLVE IT
  • keep animals away from stress, pain, discomfort, fear.
  • Keep in big cages, clean, enough food/water.
  • If have to subject to other conditions then have to be able to justify it.
25 of 40

Debriefing

  • WHAT IT IS- at end of investigation, particpants aware taken part in study, researcher discuss study with them, involve informing nature/outcome of research.
  • HOW TO SOLVE IT
  • make sure to debrief the people involved and answer questions they may have.
  • Still have right to withdraw information if uneasy.
26 of 40

Variables

  • It is assumed that chanes in the dependent variable will result after exposure to the independent variable.
  • The independent variable and potentially unwanted variables (extraneous variables) are controlled.
  • operationalised definitions of variables- specific obserable and measurable critera used to identify an abstract variable so it can be studied systematically.
  • Operationalising the IV and DV usually narrows down research focus.

Example

  • IV- drinking alcohol
  • DV- driving performance
  • Operationalised:
  • IV- 2pints beer 20mins before a driving test.
  • DV- score on "are you road ready?" simulator.
  • Population- drivers aged 18-25 in London
27 of 40

Pilot Studies

  • small-scale trial run of specific research investigation to test planned procedures and identify flaws, areas improvement, before time/money invested in main study.
  • Small no. participants used, identify problems with design, clarity of instructions, measuring instruments used, include behavioural categories.
  • Pilot questionnaire from target population.
  • Enable researcher practise research task, useful for timings, schedule on actualy day, usually inform take part pilot, so participants highlight problems in trial run.
  • Resarcher may make changes from feedback, issues raised before main study conduced.
28 of 40

Formulating Hypothesis

  • Hypothesis- a general prediction of what the results of an experiment will be, it is testable.
  • Alternate Hypothesis- predicts something other than chance variation played a part in producing results obtained. e.g. the hypothesis may predict a difference in memory recall between an experiment condition who listen to background music and a control who don't.
  • Directional- Predicts the direction in which the results are expected to occur, e.g. more words will be recalled when listening to background music than those who don't.
  • Non-directional- predicts a change but not the expected outcome, e.g. there will be a difference in the number of words recalled between those who listen to background music and those who don't.
  • Null hypothesis- predicts results are produced by chance variation alone rather than by manipulation of the IV, predicts the alternative hypothesis is not true. e.g. those who listen to background music will not recall more words than those who don't.
29 of 40

Experimental Design

Essential for success in experimental investigation:

  • provide overall plan for experiment
  • ensure appropriate/precise measurments be made
  • all data analyse correctly
  • eliminate potential sources ambiguity/bias
  • high levels control over variables

There are three main designs:

  • Independent groups design
  • Repeated measures design
  • Matched pairs design
30 of 40

Independent Groups Design

  • WHAT IT IS- involve a control and one or more experimentail conditions, participants are allocated randomly to one condition and complete the research task in only that condition.
  • Random allocation aim avoid sub/conscious bias in allocation to groups.
  • When used and is sufficient participants in each group, unlikely individual differences between groups be a confounding factor.
  • ADVANTAGES
  • No order effects occur- when performance affected by doing 2 or more experimental conditions e.g. performance in subsequent condition improved by practice of task in other attempts.
  • range potential uses when problems with order effects make repeated measures impractical.
  • DISADVANTAGES
  • error resulting from individual differences between groups of participants taking part in different conditions.
  • participants in short supply- uneconomic use as 2x as many needed to collect same amount of data as repeated measures design.
31 of 40

Repeated Measures Design

  • WHAT IT IS- participants used as control and experimental groups, so complete research task more than once under different conditions.
  • Counterbalancing- equal no. participants undertake required tasks different orders, need even no. of participants if implemented full.
  • Sometimes not appropriate apply counterbalancing to minimise order effects, problems e.g. performing one condition help performance in other more than other way round.
  • ADVANTAGES
  • individual differences removed as potential confounding variable.
  • fewer participants required, as data for all conditions collected from same group.
  • DISADVANTAGES
  • range of uses is smaller than independent groups design.
  • order effects could confound results.
32 of 40

Matched Pairs Design

  • WHAT IT IS- match each participant in one of experimental conditions as closely as possible with another participants in second condition on all variables relevant to peformance.
  • Pairs allocated then each pair randomly selected to be in each group.
  • Assumption is made that pairs can be treated as the same person so participants do only one condition, with their pair doing the other half, eliminating order effects.
  • ADVANTAGES
  • It combines the advantages of indepentent and repeated measures design.
  • DISADVANTAGES
  • difficult to get matched pairs, time consuming, depends on use reliable and valid procedures for pre-testing participants to identify matched pairs.
  • Complete matching of participants on all variables that may affect performance rarely achieved.
  • relatively uncommon, use restritcted specific situations where matching highly desirable order that experimental success can be achieved.
33 of 40

Extraneous Variables

Extra control for extra/unwanted variables, researchers want to avoid at all costs:

  • Variable other than IV producing change in DV, results are CONFOUNDED, study then useless so AVOID AT ALL COSTS.
  • Random Errors- difficult to predict e.g background noise, fatigue, hunger, thirst, state of mind.
  • Can be balanced out by randomly allocating participants randomly to conditions.
  • Constant Errors- easier to predict, issues with design e.g. time of day, study subject, educational background, intelligence level, time taken, individual differences.
  • They are more of a serious problem.
34 of 40

Reliability

  • Reliability- dependability/consistency, if the findings are replicated consistently, then the outcome is said to be reliable.
  • Observer Reliability- extent to which different observers achieve similar results when observing/recording the same participant, ideally use more than one observer in studies.
  • Inter-rater Reliability- (inter-observer), procedure used measure observer reliablity, correlational analysis used establish inter-rater reliability.

Observer reliability is achieved if there is a significant positive correlation between the scores of different observers, researchers can also use audiotapes/videotapes to check reliability of individual observers.

  • Improving observer reliability- observers trained thoroughly in techniques required, operational definitions of key terms should be clearly outlined and fully understood.
  • Test reliability- assessing the reliability of tests employed by pscyhologists, there are several methods including Split-half method and test retest method.
35 of 40

Test Reliability

  • Split half method- assess extent which individual items in particular test or questionnaire are consistent with other items in same test.
  • Split test into 2 halves after date obtained from participants, compare from odd-even no. questions or first/second half or random splitting of questions.
  • 2 sets responses correlated, significant positive is reliable, not high, researcher check procedure used, aim revision produce improvement in reliability.
  • Test retest method- stability of test or questionnaire over time, present same participants same test on 2 occasions, no feedback after first presnentation.
  • Have interval, need be selected carefully, too short participants may remember, too long, participants may have changed in meantime in way relevant to test.
  • correlational techniques used indicate test stability.
  • Significant positive correlation from sources, deemed stable, test items/procedures need to be revised if high correlation not obtained.
36 of 40

Validity

  • Internal Validity- extent be sure research findings due to mechanisms suggested, sure differences in results due to IV and not an extraneous variable.
  • External Valitidy- extend results generalised other settings beyond those which study was carried out in.
  • Population Validity- extent research findings can be generalised to other groups people.
  • Ecological Validity- extent reseach findings generalised situations outside research settings.

Test Validity

  • Face validity- simplest, independent expert assess measuring instrument is appropriate.
  • Content validity- independent expert assess validity measuring instrument, procedure rigorous, detailed exam of all parts being assessed.
  • Concurrent Validity- 2 scores same time, new procedure unknown validity and alternate procedure validity already exist correlated to assess validity, positive mean valid.
  • Predictive Validity- 2 scores obtained, different times, test high predictive validity, accurate results made about future behaviour, if test indicate certain behaviours occur but don't has low validity.
37 of 40

Bias

  • Demand Characteristics- participants try make sense of research and therefore alter behaviour include: try guess study and be un/helpful, or act out of character when observed.
  • Social Desirability Bias- participants alter behaviour as want others to see them favourably.

Investigator Effects- researchers behaviour and characteristics affect results.

  • Expectation effects- researcher want particular outcome, especially problem when observing events can be interpreted in more than one way.
  • Prescence of observer- naturalistic observation cause participants behave differently than normal behaviour.
  • Investigator characteristics- investigator age, gender, ethnicity, apperance, communication style etc. influence participants behaviour, especially true for interviews and questionnaires.
38 of 40

Measures of Dispersion

  • Measures of dispersion- tell researcher how much variability there is in the data, as measues of central tendency not always show true picture of data.
  • Standard Deviation- average distance from mean to data points.
  • Range- difference between highest and lowest scores.
  • Content Analysis- systematic research technique analysing transcripts of interviews, documents etc, coding system predetermined categories used to select materials.
  • Graphs- another way of describing and presenting patterns in quantative data.
  • Bar Charts- useful comparing categories or groups of data, consist of vertical bars equal width on an axis with varying height. Bars separated as not continous data.
  • Scatter Graphs- show correlational data, data from one variable on x axis and data from the other variable on y axis.
39 of 40

Measures of Central Tendency

  • Measures of central tendency- tells a researcher where the middle of a set of data is.
  • Mean- average score, sum of all scores divided by the number of scores.
  • Median- middle score after scores have been placed in order.
  • Mode- most frequent or most common score.
  • Skew- if the data is positively skewed the mean is greater than the mode, but if negatively skewed the mode is greater than the median
  • If data contains extreme scores that distort it, the means can be misleading.
  • The median is not affected by extreme scores and can be easily located as the middle score.
40 of 40

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »