G541: Psychological InvestigationsThe Psychological Investigations unit consists of four key approaches to psychological research: Experiments Correlations Observations Self-reportsIn addition to these are a number of other areas that fall under the heading of “PsychologicalInvestigations”: Types of experiments Research Design Sampling Methods Variables (and operationalising them) Writing hypotheses Data analysis (descriptive and inferential statistics Data (qualitative and quantitative) Types of data (nominal, ordinal, interval/ratio) Ecological validity Reliability Validity EthicsFor the exam it is important that you know and understand these issues as well as the types of researchused by psychologistsThe point of this booklet is to take you through the definitions, strengths and weaknesses of a variety ofdifferent research methods. Use the checkboxes to tick off each aspect as you go through.At the end of the booklet there is a more specific checklist relevant to the things that you should know forthe exam.
ExperimentsWhen beginning research, psychologists must decide on the type of experiment and the research designthat they wish to use in order to investigate their chosen field of study.Complete the table below, including a definition of the types of experiment.Type of Experiment Strengths WeaknessesLaboratory Experiment: Thistype of experiment isconducted in a well-controlled environmentThey allow cause and effectrelationships to beestablished.They allow for precise controlof extraneous andindependent variables.It is easier to replicate (i.e.copy) a laboratoryexperiment.Demand characteristics may bias theresults and become confounding variables.The artificiality of the setting may produceunnatural behaviour that does not reflectreal life, and results cannot be generalisedto the population.Field Experiment: Fieldexperiments are done in theeveryday (i.e. natural)environment of theparticipants but the situationsare still artificially set up. Theexperimenter stillmanipulates the IV, but in areal-life setting (so cannotreally control extraneousvariables),There is less likelihood ofdemand characteristicsaffecting the results, asparticipants may not knowthey are being studied.Behaviour in a fieldexperiment is more likely toreflect life real because of itnatural setting, i.e. higherecological validity than a labexperiment.They may be more expensive and timeconsuming than lab experiments.There is less control over extraneousvariables that might bias the results. This inturn makes the experiment harder toreplicate.Quasi/Natural Experiment:Natural experiments areconducted in the everyday(i.e. natural) environment ofthe participants but here theexperimenter has no controlover the IV as it occursnaturally in real lifeCan be used in situations inwhich it would be ethicallyunacceptable to manipulatethe independent variable.There is less likelihood ofdemand characteristicsaffecting the results, asparticipants may not knowthey are being studied.Behaviour in a naturalexperiment is more likely toreflect life real because of itnatural setting, i.e. very highecological validity.They may be more expensive and timeconsuming than lab experiments.There is less control over extraneousvariables that might bias the results. This inturn makes the experiment harder toreplicate.Complete the table below, including a definition of the types of research design.Independent measures: Different participants are used in each condition of the independent variable. Thismeans that each condition of the experiment includes a different group of participants. This should be doneby random allocation, which ensures that each participant has an equal chance of being assigned to onegroup or the other.Repeated Measures: The same participants take part in each condition of the independent variable. Thismeans that each condition of the experiment includes the same group of participants.Matched Pairs:Which allows similar participants to be matched against each other in different conditions.AMPD is when a sample are divided into different conditions or groups based on their characteristics. Thisidea being to try and create two groups that possess a similar a set of “features” as possible without thembeing the same people. This is often done either by directly matching one person to another or by quotesampling, depending on the characteristic that you want to keep the same.
Research Design Strengths WeaknessesIndependent Measures Avoids order effects (such aspractice or fatigue) as peopleparticipate in one conditiononly. If a person is involvedin several conditions theyman become bored, tiredand fed up by the time theycome to the secondcondition, or becoming wiseto the requirements of theexperiment!Each participant comes freshand un-practised to eachcondition. If they wererepeated by another similarset of participants, similarresults would be found. Thisgives the study greaterreliability.There is greater validity asthey can be sure that theeffects of the IV aremeasured rather than theeffect of boredom, practice ordemand characteristicsMore people are needed than with therepeated measures design (i.e. more timeconsuming)Differences between participants in thegroups may affect results, for example;variations in age, sex or social background.These differences are known as participantvariables (i.e. a type ofextraneousvariable).The scores between participants in thedifferent conditions can be less reliablycompared as differences in scores may bedue to differences between participantsrather than the effect of the IV.The results are less likely to be valid asindividual differences can have an effect onthe DV. We cannot be certain that what ismeasured is caused by manipulation of theIVRepeated Measures Fewer people are needed asthey take part in allconditions (i.e. saves time)More control over individualdifferencesThe scores betweenparticipants in the differentconditions can be reliablycompared as differences inscores will be due to theeffect of the IV as there areno differences betweenparticipants who take part inboth conditions.The results are likely to bemore valid as what ismeasured is more likely to bethe effect on the IV as thereare no individual differencesto affect the result.There may be order effects. Order effectsrefer to the order of the conditions havingan effect on the participants‟ behaviour.Performance in the second condition maybe better because the participants knowwhat to do (i.e. practice effect). Or thereperformance might be worse in the secondcondition because they are tired (i.e.fatigue effect)The results may be less reliable becausethe participants may have improved withpractice. This may make the differencebetween the two conditions less reliable asdifferent results could be obtained simplyby changing the order of the conditions.Taking part in more than one conditionreduces validity as participants have agreater chance of working out the aim ofthe study. The results are thereby affectedby demand characteristics.Matched Pairs More control over individualdifferencesThe scores betweenparticipants in the differentconditions can be reliablycompared as differences inscores will be due to theeffect of the IV as relevantdifferences betweenparticipants are matchedover the conditions.The results are likely to bemore valid as what ismeasured is more likely to beCosts of bringing participants in.Impossible to find exactly the samematches.Difficult to organise so only allows for smallsamples.
the effect on the IV asrelevant individualdifferences are matched overthe conditions to have less ofan impact on the result.Each participant comes freshand un-practised to eachcondition. If they wererepeated by another similarset of participants, similarresults would be found. Thisgives the study greaterreliability.There is greater validity asthey can be sure that theeffects of the IV aremeasured rather than theeffect of boredom, practice ordemand characteristics.Once this has been decided, researchers must manipulate and measure variables. There are differenttypes of variables, including those that affect the experiment and cause misleading results:Variables DefinitionIndependent Variable The IV is the variable that changes between conditions. Sometimes called theexperimental variable.Dependent Variable The DV is the variable that is measured by the experimenter. This will make upthe results of the experiment.Confounding Variable Confounding variables are those that we are not aware of or controlling but theywork against our hypothesis and suggest that there is no relationship betweenour IV and V.Extraneous Variable 1:Participant VariablesExtraneous variables are those which we have not taken into account, are notcontrolling or even may not be aware of their existence.Extraneous Variable 2:Situational VariablesIf we don‟t do all we can to eradicate them, we are opening up the possibility thatour results will not be an accurate reflection of the relationship between our IVand DV. Something else may be at work and thus we will draw false conclusionsabout why a relationship exists.Situational VariablesAspects of the research situation other than the IV which mayinfluence the DV.For example, if you were doing an experiment and there weresome workers outside making a lot of noise which might distractthe participants.Experimenter VariablesEffects of the experimenter‟s expectations which arecommunicated intentionally or unintentionally.For example, an experimenter might accidentally give you clues asto the „right‟ answer in a test by their use of body language or theirtone of voice.Participant Variables Aspects of the participant‟s characteristics or experience (other
than the IV) which might influence the DVFor example, you might be looking at reaction times and justhappen to have some participants who are particularly good at fastmoving sports like tennis, and who therefore might have generallyfaster reaction times.Controls are designed to reduce or remove the effects of extraneous variables. Controls make theresearch more valid as you can be more certain that it really is only the IV affecting the DV. Controls canalso make the research more reliable as it usually means you have a standardized procedure.To remove situational variables the researcher might try to control the environment. For example, makingsure that all participants complete the same task under exactly the same conditions. Many psychologistsuse small, sound-proof rooms to carry out their experiments so that participants cannot be influenced byoutside distractions and every participant has exactly the same experience.To control for participant variables the researcher might consider which experimental design they want touse. A repeated measures design rules out participant variables as all participants do the same test underdifferent conditions of the IV. Alternatively, a matched pairs design means that participants have beenmatched up on specific characteristics (eg IQ, age, personality) so that the only differences between groupsare those that have been deliberately manipulated by the researcher.Following decisions made on the type of experiment and research design, the experimenter must makepredictions called hypotheses. The statement of the hypothesis must contain the independent anddependent variables and they must be clearly operationalised so that the hypothesis details what is beingmeasured and how.A researcher will start with a research question and then write out specific predictionsbased on it.Using the research question, “Does background noise affect our ability to remember?”complete the table below including a relevant hypothesis.Hypotheses Definitions ExampleAlternateHypothesisThe experimental(alternative) hypothesis candirectional or non-directionalCan be one tailed or two tailedNull Hypothesis It is predicting that norelationship or link will befound between the twoconditions or the variablebeing tested.“Those participants in the smokingcondition (IV) will report a significantlymore fearful attitude towards risk (DV) ona Likert scale rating than those in the nonsmoking condition.”One-TailedHypothesisSometimes a hypothesispredicts the direction in whichthe results are expected togo. A one-tailed hypothesisstates the direction in whichthe results are expected toFor example, “studying significantlyimproves exam marks”
go (i.e. in one direction). E.g.“boys will recall more wordsthan girls”.Two-TailedHypothesisIf a hypothesis does not statea direction but simply statesthat one factor will affectanother, or that there will bea difference between twosets of scores (without sayingwhich direction thatdifference will be) then it isknown as a two-tailed (ornon-directional) hypothesis.oA two-tailedhypothesis does not give apredicted direction. E.g.“there will be a difference inthe number of words girlsand boys recall”.For example, “anxiety influencesperformance”Key words:Key words DefinitionsEcological validity This type of validity refers to how well a study can be related to or reflects every day,real life. Studies with high ecological validity can be generalised beyond the settingthey were carried out in, whereas studies low in ecological validity cannot.Reliability Reliability refers to the extent to which a measure is consistent within itself.Validity Validity refers to whether a study measures or examines what it claims to measure orexamine.DemandcharacteristicsA feature of an experiment (other than the IV) that influences a participant to try toguess what a study is about and look for clues as to how to behave. It is any aspectof a study which has an influence on participants to do or answer what is expected ofthem.Counterbalancing This is a strategy often used when carrying out a repeated measures design tocontrol for potentially confounding variables (i.e. order effects). Thereforecounterbalancing involves alternating the order in which participants do theconditions of the experiment.For example a half of the participants could completecondition A and then condition B and the other half could complete condition B andthen condition A.Order effects When carrying out a repeated measures design it is possible that order effects suchas practice effect or fatigue effect could produce an unwanted effect.Words to be added:CorrelationsCorrelation refers to a measure of how strongly two or more variables are related to each other.Figures DefinitionsREALCONSISTENTMEASURESGUESSINFLUENCESEXPECTEDREPEATED ALTERNATINGFATIGUE
A positive correlation means that high values of one variable are associatedwith high values of the other. Or if you like, the variables increase together.Fig 1. Positive correlation; for example as students’ optimism increases, sodo their gradesA negative correlation means that high values of one variable areassociated with low values of the otheror as one variable increases theother decreases. Note that like a positive correlation, a negative correlationstill indicates that some kind of relationship exists.Fig 2. Negative correlation; for example as stressful life experiencesincrease, students’ grades decreaseIf there is no correlation between two variables they are said to beuncorrelated.Fig 3. No correlation; for example, no relationship between students gradesand if they own a pencil caseIt is important to remember that association does not mean causation. For example, there is almostcertainly a very high positive correlation between the length of people‟s right arm and the length of their leftarm. But the length of a person‟s left arm did not determine the length of their right arm. They are bothdetermined by other factors i.e. genetics, diet etc.A correlation coefficient refers to a number between -1 and +1 and states how strong a correlation is. Ifthe number is close to +1 then there is a positive correlation. If the number is close to -1 then there is anegative correlation. If the number is close to 0 then the variables are uncorrelated.Hypotheses for correlationsCorrelational hypotheses predict a relationship between variables, not a difference or effect.Complete the table for the following hypotheses using the research question “Will there be a relationshipbetween people who perceive their memory to be good and their actual memory ability?”Make sure that you write it with regards to a relationship between the variables rather than a difference.Strengths of correlations1. Correlation allows the researcher to investigate naturally occurring variables that maybe unethical orimpractical to test experimentally. For example, it would be unethical to conduct an experiment on whethersmoking causes lung cancer.2. Correlation allows the researcher to clearly and easily see if there is a relationship between variables.This can then be displayed in a graphical form.Limitations of correlations1. Correlation is not and cannot be taken to imply causation. Even if there is a very strong associationbetween two variables we cannot assume that one causes the other. For example suppose we found apositive correlation between watching violence on T.V. and violent behaviour in adolescence. It could bethat the cause of both these is a third (extraneous) variable - say for example, growing up in a violent home- and that both the watching of T.V. and the violent behaviour are the outcome of this.
2. Correlation does not allow us to go beyond the data that is given. For example suppose it was found thatthere was an association between time spent on homework (1/2 hour to 3 hours) and number of G.C.S.E.passes (1 to 6). It would not be legitimate to infer from this that spending 6 hours on homework would belikely to generate 12 G.C.S.E. passes.Hypotheses ExampleAlternate hypothesis:The experimental(alternative) hypothesiscan directional or non-directionalSometimes a hypothesis predicts the direction in which the results are expectedto go. For example, “studying significantly improves exam marks”; “Women aresignificantly better drivers than men”.Null hypothesis: nodirectionThere will be no significant correlation between the amount of alcohol consumedand scores on a memory testOne-tailed hypothesis:When hypotheses predictthe direction of theresultsA one-tailed hypothesis states the direction in which the results are expected togo (i.e. in one direction). E.g. “boys will recall more words than girls”.Two-tailed hypothesis If a hypothesis does not state a direction but simply states that one factor willaffect another, or that there will be a difference between two sets of scores(without saying which direction that difference will be) then it is known as a two-tailed (or non-directional) hypothesis. For example, “anxiety influencesperformance”; “there is a significant difference in the driving performance of menand women”.Self-ReportsA self report is any method which involves asking a participant about their feelings attitudes and so on.Examples of self reports are questionnaires, interviews but note that self reports are often used as a way ofgaining participants responses in observational studies and experiments.The two main types of self-reports are questionnaires and interviews. Either of these can be used in orderto collect qualitative and/or quantitative data. Complete the table below including definitions, strengths andweaknesses for quantitative and qualitative data.Data Type Strengths WeaknessesQuantitative Easier to analyseProduces neat conclusionsOversimplifies reality and humanexperience – statisticallysignificant but not humanelysignificantQualitative Represents true complexities ofhuman behaviourGains access to thoughts andfeelings which may not beassessed using quantitativemethods with closed questionsProvides rich detailMore difficult to detect patternsand draw conclusionsSubjective analysis can beaffected by personal expectationsand beliefsQuestionnaires can use a variety of different question types and they can be handed out, emailed/posted,face to face or phone. Complete the definitions, strengths and weaknesses below:Question Types Strengths WeaknessesOpen questions •Provide a rich source of •Harder to analyse and make
qualitative information (i.e. moredescriptive than numerical) asthere is no restriction to theresponse and therefore theiranswers are valid for them.•Less researcher led due to theparticipant being in control of theirown responsecomparisons and correlationsfrom.•Poor reliability due to individualinterpretations from thoseanswering questions.Closed questions •The data from closedquestionnaires are very easy tocollect and summarise due to thesimplicity of response required.•Data is often quantitative (i.e.numerical) so comparison iseasier.•Very reliable as answers willalways be in the same format andeasy to administer.•Potentially valid due to likelihoodof participant understandingexactly what is required of themwithout involvement form theresearcher.•Lacks the richness of openended questions = less validity•Researcher led – you‟ve decidedthe answers and the question =may lead to missing vital data thathasn‟t been previouslyconsidered.•Less answers possible, lesschoice for participant = less validfor them•Only useful for simple issues.•Not appropriate for ethicallysensitive topics.Questionnaires use other specific types of questioning techniques. For example they will use rating scalesin order to obtain quantitative data that expresses degrees of opinion or forced choice questions. Use thetable below to complete the details:Questioning Types (define withan example)Strengths WeaknessesRating Scales Likert Scale is a five point scalewhich is used to allow theindividual to express how muchthey agree or disagree with aparticular statement.•Likert Scales have the advantagethat they do not expect a simpleyes / no answer from therespondent, but rather allow fordegrees of opinion and even noopinion at all. Thereforequantitative data is obtained,which means that the data can beanalysed with relative ease.•You can summarise largenumbers of peoples‟ attitudesthrough mode and mediananalysis.•High control over answers =highly reliable and easy forresponder to understand•Small number of answers = easycomparison between answers todraw out correlations, patternsand trends•However, like all surveys, thevalidity of Likert Scale attitudemeasurement can becompromised due the socialdesirability. This means thatindividuals may lie to putthemselves in a positive light. Forexample, if a Likert scale wasmeasuring discrimination, whowould admit to being racist?•You can really look at the meanof the data. You have to startputting values on answers whichwere not there when theparticipant agreed to them.•Not enough answers to choosefrom - lowers validity.Forced Choice
Interviews on the other hand will nearly always be conducted face to face. There are also different types ofinterview (see below – you know what to do!)Interviews are a type of spoken questionnaire where theinterviewer records the responses. Interviews can be structured whereby there is a predetermined set ofquestions or unstructured whereby no questions are decided in advance.Interview Types Strengths WeaknessesStructured interviews: pre-setquestions asked in same orderwithout variationFavoured by quantitativeresearchers, easier to replicateand compare results, less chanceof interviewer biasNo opportunity for probingdeeper, less chance ofdiscovering new hypothesis,harder to discover what isimportant to the respondentsUnstructured interviews: few or nofixed questions, more like aconversationFavoured by qualitativeresearchers, allow respondants todirect the interview, moreopportunity for developing newhypothesisHarder to replicate, time-consuming, difficult to compareinterviews, may go off trackInterviews are a type of spoken questionnaire where the interviewer records the responses. Interviewscan be structured whereby there is a predetermined set of questions or unstructured whereby no questionsare decided in advance.The main strength of self-report methods is that they are allowing participants to describe their ownexperiences rather than inferring this from observing participants.Questionnaires and interviews are often able to study large samples of people fairly easy and quickly.They are able to examine a large number of variables and can ask people to reveal behaviour and feelingswhich have been experienced in real situations.However participants may not respond truthfully, either because they cannot remember or because theywish to present themselves in a socially acceptable manner. Social desirability bias can be a bigproblem with self report measures as participants often answer in a way to portray themselves in a goodlight.Questions are not always clear and we do not know if the respondent has really understood the questionwe would not be collecting valid data.Unstructured interviews can be very time consuming and difficult to carry out whereas structuredinterviews can restrict the respondents‟ replies. Therefore psychologists often carry out semi-structuredinterviews which consist of some pre-determined questions and followed up with further questions whichallow the respondent to develop their answers.Other weaknesses (write them in the context of a self-report related weaknessA self report is any method which involves asking a participant about their feelings, attitudes, beliefs and soon. Examples of self reports are questionnaires and interviews but note that self reports are often used as away of gaining participants responses in observational studies and experiments.
The main strength of self-report methods are that they are allowing participants to describe their ownexperiences rather than inferring this from observing participants.Questionnaires and interviews are often able to study large samples of people fairly easy and quickly. Theyare able to examine a large number of variables and can ask people to reveal behaviour and feelings whichhave been experienced in real situations.However participants may not respond truthfully, either because they cannot remember or because theywish to present themselves in a socially acceptable manner. Social desirability bias can be a big problemwith self report measures as participants often answer in a way to portray themselves in a good light.Questions are not always clear and we do not know if the respondent has really understood the questionwe would not be collecting valid data.If questionnaires are send out, say via email or through tutor groups, response rate can be very low.Questions can often be leading. That is, they may be unwittingly forcing the respondent to give a particularreply.Unstructured interviews can be very time consuming and difficult to carry out whereas structured interviewscan restrict the respondents‟ replies.Therefore psychologists often carry out semi-structured interviews which consist of some pre-determinedquestions and followed up with further questions which allow the respondent to develop their answers.Social desirabilityDemand characteristics People will respond change their behaviour depending on what they think isexpected of them. Just because they know that they are being studied or undersome form of scrutiny will mean that their behaviour will not be natural.This means that any situation where someone knows that they are beingstudied could be said to produce “demand characteristics” and this means thattheir answers, actions, responses and performances may not be valid.Response ratesLeading questions Fixed choice questions are phrased so that the respondent has to make a fixedchoice answer usually „yes‟ or „no‟.One of the most common rating scales is the Likert scale. A statement is used and the participantdecides how strongly they agree or disagree with the statements. For example the participantdecides whether they strongly agree/ agree/ undecided/ disagree/ strongly disagree that Mozzarellacheese is great.A strength of Likert type scales is that they can give us an idea about how strongly a participantfeels about something. This therefore gives more detail than a simple yes no answer.A further strength is that the data are quantitative data which are easy to analyse statistically.However there is a tendency with Likert scales for people to respond towards the middle of thescale perhaps to make them look less extreme.As with any questionnaire participants may provide the answers that they feel they should andimportantly as the data is quantitative it does not provide in depth replies.
ObservationsThere are four different types of observation. Complete the table with strengths and weaknesses.Observation Types (plusdefinitions)Strengths WeaknessesParticipant observation:Participant observation is whenthe researcher joins in with thegroup he/she is studying to get adeeper insight into their lives.Observational studies areinvestigations where theresearcher observes a situationand records what happens butdoes not manipulate anindependent variable.Observational studies thereforetend to be high in ecologicalvalidity as there is no interventionand if the observer remainsundetected the method avoidsproblems with demandcharacteristics.A main strength of observationalstudies is that they get to see howparticipants actually behaverather than what they say they do.A further strength of observationalstudies is that they offer ways ofstudying behaviour when thereare ethical problems withmanipulating variables.On the other hand observationalstudies are difficult to replicate.Observations do not provideinformation about whatparticipants are thinking orfeeling.There is little or no control ofextraneous variables inobservational studies thereforewe can not make cause and effectstatements.There is also the problem ofobserver bias with observationalstudies. This occurs if theobservers „see‟ what they expectto see.A number of ethical issues canarise with observational studiesincluding problems with a lack ofinformed consent and invasion ofprivacy.Observations can also be verytime consuming, require carefulpreparation and possibly trainingfor the observers.Non-participant observation: Anon-participant observationis a type of observational studywhereby the researcher does notjoin in with the activity beingobserved.Saves time and money andrequire less confederates, canuse a coding schemeParticipants may see that peopleare observing their behaviours somay act differently, i.e. so may actdifferently due to demandcharacteristics or sociallydesirable behaviours.Disclosed (overt) observation:when the people being observeddo not know theyre beingwatched or studiedGains qualitative dataThe avoidance of problems ofethics in that the group are awareof the researchers role.the group is being observed in itsnatural setting.Data may also be openlyrecorded.Problems of going native areavoided.UnethicalTime consumingCan be costly to spend all thistime observing in comparison toemailing people questionnairesfor example.Observer effect, where thebehaviour of those under studymay alter due to the presence ofthe researcher.Undisclosed (covert) observation:when they have been told andgiven consent to be observed bythe researcherEthically sound, gains qualitativedataThe social researcherparticipating fully withoutinforming members of the socialgroup of the reasons for herpresence, thus the research iscarried out secretly or covertly.contact with a gatekeeper, amember of the group under studywho will introduce the researcherinto the group.the researcher may gain accessCan cause the hawthorne effect(because they know theyre beingobserved)Time consumingCan be costly to spend all thistime observing in comparison toemailing people questionnairesfor example.the researcher having to becomeinvolved in criminal or dangerousactivities, particularly where theresearch is studying a deviantsocial group.
to social groups who wouldotherwise not consent to beingstudied.The avoidance of problems ofobserver effect, the conceptionthat individuals behaviour maychange if they know they arebeing studied. However, there areproblems of recording data.problems of negotiating andhaving to act out forms ofbehaviour which the researchermay personally find unethical ordistasteful.the researcher having to employ alevel of deceit, since theresearcher is essentially liesabout the nature of her presencewithin the group.close friendships are oftenresulting from connections withmembers of the group understudy and the covert nature of theresearch can put a tremendousstrain on the researcher, both inand out of the fieldwork setting.the problem of going native,which refers to the fact that aresearcher will cease to be aresearcher and will become a full-time group participant.A structured observation is where the researchers design a type of coding scheme to record theparticipants behaviour.Structured observations generally provide quantitative data. Coding schemes are ways of categorisingbehaviour so that you can code what you observe in terms of how often a type of behaviour appears.Advantages of using a coding scheme are that they are fairly simple to carry out and that they providequantitative data which can be analysed statistically.However observation using coding schemes has a main weakness. It gives a very restricted view of what isactually happening. The researcher may miss important behaviour and the data is not as in-depth as simplyobserving behaviour which is occurring.An unstructured observation involves the researchers recording the behaviour they can see. This can bedifficult without the use of recording equipment (such as a video camera), can be difficult to analyse butdoes provide rich qualitative data.Furthermore with unstructured observations there is a tendency for observers to record the most eye-catching or noticeable behaviours which might not be the most relevant or important behaviours to recordObservations use two different types of sampling for recording behaviours.Sampling Types Strengths WeaknessesTime Sampling:Time sampling refers toresearchers choosing timeintervals for making observationseither systematically or randomly.The frequency and types ofbehaviour are then recorded for aset time.Allows you to take a sample atthat particular time to recorddesired behavioursBehaviours at that time may bedifferent to behaviours at anothertime so is therefore situationallyaffected.Event Sampling: Whenresearchers are interested inAllows for inter-rater reliability ifyou are using more than oneParticipant may see personrecording so may be subjected to
events that happen infrequently,they rely on event sampling tosample behaviour. It may betherefore, that you watch thewhole game of football to see howmany times red cards are given inmatches.confederate to observebehavioursdemand characteristics andsocially desirable behaviours orbias. Subjected to individualdifferences as well as situationalfactors.Time sampling involves observations at set lengths of time at set intervals (e.g. in a traffic survey 3 hourlyobservations between 08.00-09.00, 12.00-13.00 and 17.00-18.00). Event sampling involves observations ofa specific event each time it occurs throughout the duration of the observation period.Observation studies are those where the researcher observes a situation and records what happens butdoes not manipulate an independent variable. Observational studies therefore tend to be high in ecologicalvalidity as there is no intervention and if the observer remains undetected the method avoids problems withdemand characteristics.Discuss the issues of reliability and validity in terms of observational research (mention observer bias andinter-rater reliability in one or the other).IssuesReliabilityReliability refers to how consistent a measure is.A procedure or test is reliable if all participants can complete the procedure in the samewayValidity Validity refers to whether a study measures or examines what it claims to measure orexamineHow can you test or improve the reliability of your research?Use a standardized procedure where each participant experiences the experiment in exactly thesame way. That way it doesn‟t matter so much exactly how you‟re measuring, just as long as it‟s thesame for everyone.Conduct a pilot study. This is a smaller scale version of the test carried out before you conduct theresearch for real. This allows you to iron out any problems with the procedure, check for reliability ofyour measures and also make sure that it is worth doing the research.Use the test-re-test method. This involves doing a test once (eg giving out a questionnaire) andthen redoing it again at a later date, e.g one week later. The results can then be compared to see ifthey are the same.
And if you are observing behaviour, use multiple observers to gain inter-rater reliability. This is theextent to which separate observers agree on what has been observed. This might involve twopeople watching the same behaviours for a fixed period of time, then comparing the results theyhave recorded. If there is over 80 % agreement that they have collected the SAME behaviours thenthey can be said to have high inter-rater reliability.There are several types of validity:Concurrent Validity: a method for assessing validity by comparing the measure to some othermeasure done at the same time: if the measure is valid the findings should match up and be thesame.Face Validity: the degree to which a test or measure appears to look as though it probablymeasures what it is supposed to.Ecological Validity: the extent to which the findings of a piece of research can be generalised to reallife.Sampling MethodsOne of the most important issues about any type of method is how representative ofthe population the results are. The population is the group of people from whom thesample is drawn.o A sample is the group of people who take part in the investigation. Thepeople who take part are referred to as “participants”.o Sampling is the process of selecting participants from the population.o The target population is the total group of individuals from which the sample might be drawn.o Generalisability refers to the extent to which we can apply the findings of our research to the targetpopulation we are interested in.Obviously it is not usually possible to test everyone in the target population so therefore psychologists usesampling techniques to choose people who are representative (typical) of the population as a whole.Generalisability refers to the extent to which results from one sample of participants can be applied to widergroups. The generalisability of the results of a study is partly dependent on the success of the samplingtechnique (e.g. was the sample representative of the population) and the representativeness of thepopulation chosen (for example if the sample was taken from students then it is not reasonable togeneralise the results to all types of people).There are four main sampling methods (define the method with strengths and weaknesses).Operationalize: how you are going to measure your independent and dependent variablesSay you want to test how good a goalkeeper is, you operationalize the experiment by counting how manytimes they make a save.When conducting experiments it is important to operationalize the variables. That is, stating a clear waythat the independent variable is going to be manipulated and the dependent variable is to be measured.
For example if an experiment was to be carried out to see if time of day affected memory it would beimportant to operationalize the variables of time of day and memory. We might operationalize time of dayas 10am and 10pm and operationalize memory as performance on a memory task.In fact, it may be worth being even clearer with the operationalizing of the dependent variable and statehow the performance on a memory task is to be measured. For example number of words recalled out of alist of fifty words.Sampling Methods Strengths WeaknessesOpportunity sample: It consists oftaking the sample from peoplewho are available at the time thestudy is carried out and fit thecriteria you are looking for.It is a popular sampling techniqueas it is easy in terms of time andtherefore money. For examplethe researcher may use friends,family or colleagues. It can alsobe seen as adequate wheninvestigating processes which arethought to work in similar ways formost individuals such as memoryprocesses. Sometimes,particularly with naturalexperiments, opportunitysampling has to be used as theresearcher has no control overwho is studied.Opportunity sampling canproduce a biased sample as it iseasy for the researcher to choosepeople from their own social andcultural group. This sample wouldtherefore not be representative ofyour target population as youfriends may have differentqualities to people in general.A further problem with opportunitysampling is that participants maydecline to take part and yoursampling technique may turn intoa self-selected sample.Random sample: Randomsamples require a way of namingor numbering the targetpopulation and then using sometype of raffle method to choosethose to make up the sample.Random samples are the bestmethod of selecting your samplefrom the population of interest.Your sample should represent thetarget population and eliminatesampling bias: Random samplingis the best technique for providingan unbiased representativesample of a target population.It is very difficult to achieve (i.e.time, effort and money).Random sampling can be verytime consuming and is oftenimpossible to carry out,particularly when you have a largetarget population, of say allstudents. For example if you donot have the names of all thepeople in your target populationyou would struggle to obtain arandom sample. If you askpeople to volunteer for a study thesample is already not random assome people may be more or lesslikely to volunteer for things.Similarly if you decided to put outan advert for participants it wouldbe almost impossible toguarantee that every member ofyour target population has anequal chance of viewing theadvert.Self-selected sample: Self-selected sampling (or volunteersampling) consists of participantsbecoming part of a study becausethey volunteer when asked or inresponse to an advert. Thissampling technique is used in anumber of the core studies, forexample Milgram (1963).This technique, like opportunitysampling, is useful as it is quickand relatively easy to do. It canalso reach a wide variety ofparticipants.However, the type of participantswho volunteer may not berepresentative of the targetpopulation for a number ofreasons. For example, they bemore obedient, more motivated totake part in studies and so on.Snowball sample: members of asample put researcher in touchwith other membersMainly used with groups who arehard to identify or accessNot very representative as it isbased only on people who havecontact with one another
Systematic Sampling: Choosesparticipants in a systematic (i.e.orderly) way from the targetpopulation, like every nthparticipant on a list of names.Totake a systematic sample, you listall the members of the population,and then decided upon a sampleyou would like. By dividing thenumber of people in thepopulation by the number ofpeople you want in your sample,you get a number we will call n. Ifyou take every nth name, you willget a systematic sample of thecorrect size. If, for example, youwanted to sample 150 childrenfrom a school of 1,500, you wouldtake every 10th name.The advantage to this method isthat is should provide arepresentative sampleIs that it is very difficult to achieve(i.e. time, effort and money).Stratified Sampling: Theresearcher identifies the differenttypes of people that make up thetarget population and works outthe proportions needed for thesample to be representative.A list is made of each variable(e.g. IQ, sex etc.) which mighthave an effect on the research.For example, if we are interestedin the money spent on books byundergraduates, then the mainsubject studied may be animportant variable. For example,students studying EnglishLiterature may spend moremoney on books than engineeringstudents so if we use a very largepercentage of English students orengineering students then ourresults will not be accurate. Wehave to work out the relativepercentage of each group at auniversity e.g. Engineering 10%,Social Sciences 15%, English20%, Sciences 25%, Languages10%, Law 5%, Medicine 15% Thesample must then contain allthese groups in the sameproportion as in the targetpopulation (university students).The sample should be highlyrepresentative of the targetpopulation and therefore we cangeneralise from the resultsobtained.Gathering such a sample wouldbe extremely time consuming anddifficult to do
TaskA researcher has been asked to complete a survey of students‟ attitudes to banning smoking in the collegegrounds.(a) Describe how the researcher might carry out an opportunity sampling technique for this study. _______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________(b) Outline one strength and one weakness of using an opportunity sampling technique for this study. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________Data and data analysisThe data that can be gathered in studies is known as quantitative and qualitative. Use the table below tocomplete the strengths and weaknesses and this time specify research methods relevant to each data type:Data types Strengths Weaknesses Research methodsQualitative Reliable or valid?Quantitative Reliable or valid?Data from research is analysed using different statistical approaches. Define them and give strengths andweaknesses of using them.A type 1 error is rejecting a null hypothesis that is true – this happens when the significance level is toohighA type 2 error is rejecting a main hypothesis that is true (and accepting the null)– this happens when thesignificance level is too low
Statistics Strengths WeaknessesDescriptive statistics: Descriptive= visualGraphsCentral tendency (mean, median,mode)Measure of dispersion (range,standard deviation)Describe how spread out thevalues in a data set areDescriptive statistics simply offerthe researcher ways of describingand summarising quantitativedata.Measures of central tendencydescribe a data set by identifyingone score that represents thegeneral trend of that data. Theydescribe how the data clustertogether around a central pointThere are three measures ofcentral tendency1.Mode2.Median3.MeanRange:The difference betweenthe highest and lowest scores in aset of dataQuick to calculateGives us a basic measure of howmuch the data variesTells us nothing about data in themiddle of a set of scoresAffected by outlying valuesHowever, other researchers arguethat extreme scores makemeasures unreliable as theymisrepresent the true tendency ofa data set by skewing it so it is toohigh or too low.However, other researchers arguethat extreme scores makemeasures unreliable as theymisrepresent the true tendency ofa data set by skewing it so it is toohigh or too low.Some researchers also dislike themean because the average isoften a decimal score, which theysay is meaningless. Meanwhile,the median is normally one of thescores in the data set and themode is always one of the scoresin the data set.The main problem with the modeis that it relies on there being ascore which occurs more thanothers. However, sometimeseach score in a data set occursonly once so there is no mode assuch. Alternatively, there may betwo or more scores that occurequally frequently, which gives anumber of averages. Sometimethese can be at either end of arange of data so do not really helpin measuring central tendency!Inferential statistics: Inferential =significanceSpearman‟s rho: Used to testwhether the correlation betweentwo co-variables is significant ornot. Not nominalChi square:observation/data intest must be actual values (e.g.total number of people who…)must not be averages,percentages or proportions.In chi-square, if a small sample isused there is a risk of a type 1error. This should be minimisedby using a larger sample.Mann Whitney: Uses dataobtained in the study. Nominal? –no (ordinal)Correlation/relationship? – no(difference)Independent groups – yes! Usedin experimentsWilcoxon T test: Repeated
measures design(can also be matched pairs)DifferenceData ordinal or intervalMiddle value of scores arranged in rank orderUnlike the MEAN, it can be used on ranked data, e.g., placing a group of people in order of position on amemory test rather than counting their actual scoreUnaffected by extremes, so can we can use it on data with a skewed distribution where results would be abit one sidedDoes not work well on small data setsNot as powerful as the MEAN: we can only say one value is higher than another on ranked dataMost frequently occurring value in a data setUnaffected by extremes as we‟re just looking at the most common value rather than its positionCan be used on basic data forming nominal categories – we could do a frequency count on these, e.g.,number of people preferring vanillaSmall changes can make a big differenceCan be bi/multimodal, e.g., 3,5,8,8,10,Standard deviation: Measures the variability of our data, i.e., how scores spread out in relation to the meanscore. Allows us to make statements about probability – how likely or unlikely a given value is to occur.Most powerful measure of dispersion as all the data is used. Data cannot be ranked or from categorie. Datamust form a normal distribution curve as SD is affected by skewed dataOnce data has been collected it can be categorised by the way in which it has been collected.Data type anddefinitionMeasure of centraltendencyStrengths WeaknessesNominal – data ispresented in categoriesChi Squared – nominallevel data
Ordinal – data is rankedin orderSpearman‟s Rho –correlations with ordinallevel dataMann Whitney U – atleast ordinal level data inan independent groupsdesign (i.e. looking for adifference between twodatasets)Interval/Ratio - Preciseand measured usingunits of equal intervalsEthicsThese are a set of guidelines which psychologists carrying out research should follow.The following box includes a summary of the ethical criteria proposed by the British Psychological Societyfor the conduct of research.The British Psychological Society issued revised ethical principles in June 1990. In the conduct of theirresearch, psychologists should always consider the following;Consent; Have the subjects of the study made an informed consent to take part? Have the parents of childsubjects given informed consent to the research procedures? Have payments been used to induce risktaking behaviour?
Deception; Have the subjects been deceived? Was there any other way to carry out the study other thanby using deception? Have the procedures been approved by other psychologists?Debriefing; Have the subjects been effectively debriefed? Has any stress caused by the procedures beenremoved?Withdrawal from the investigation; Are the subjects clear that they can withdraw from the study at anytime without penalty or scorn?Confidentiality; Participants in psychological research have the right to expect that information theyprovide will be treated confidentially.Protection of participants; Investigators must protect participants from physical and mental harm duringthe investigation.Observational research; Unless the participants give the consent to being observed, observationalresearch must only take place where those observed could normally be expected to be observed bystrangers.Giving advice; Psychological advice must only be given if the psychologist is qualified in the area that theadvice is requested in.Colleagues; Psychologists should take action if they believe that any of the above principles are beingviolated by a colleague.Identify a core study (or studies) that broke the following ethical guidelines and how:Guideline Core study (or studies) and howInformed consentRight to withdrawProtection of participantsDeceptionExperimental Tick SheetStuff you must know for experiments Identify two general strengths and two weaknesses of experiments Describe one strength and one weakness of laboratory experiments Describe one strength and one weakness of field experiments Identify and describe an independent measures design Know one strength and one weakness of an independent measures design Identify and describe a repeated measures design Know one strength and one weakness of an independent measures design
Describe and evaluate an appropriate procedure Write a one-tailed experimental hypothesis (alternate) Write a two-tailed experimental hypothesis (alternate) Write a null hypothesis for an experiment Identify the independent and dependent variables in an experiment Suggest how independent and dependent variables can be operationalised An alternative way to operationalise the dependent variable A strength and weakness of the alternative way of operationalising the dependent variable. Know how to calculate a mean, median and mode Know how to work out the range and what this means Identify one reason why the mean is a useful measure of central tendency Identify one reason why the median is a useful measure of central tendency Identify one reason why the mode is a useful measure of central tendency Sketch an appropriate and labelled bar chart of the mode, median or mean Draw conclusions from a table Draw conclusions from a bar chart Ethical Issues associated with the use of experimentsSelf-report tick sheetStuff you must know for self-report measures Identify two general strengths and two weaknesses of self reports Describe a strength and weakness of rating scales Describe a strength and weakness of closed questions Describe a strength and weakness of open questions Know what is meant by reliability Know how to test a self-report for reliability Know how to improve the reliability of a self report Know what is meant by validity Know how to test a self-report for validity Know how to improve the validity of a self report Describe and evaluate an appropriate procedure Describe what is meant by quantitative data Outline one strength and one weakness of collecting quantitative data using self-report techniques Describe what is meant by qualitative data Outline one strength and one weakness of collecting qualitative data using self report techniques Ethical issues relating to self report techniquesStuff you must know for sampling methods Describe opportunity sampling
Identify two strengths and two weaknesses of opportunity sampling Describe random sampling Identify two strengths and two weaknesses of random sampling Describe self-selected sampling Identify two strengths and two weaknesses of self-selected samplingCorrelation tick sheetStuff you must know for correlations Identify two general strengths and two weaknesses of correlations Know what is meant by a positive correlation Know what is meant by a negative correlation Write a one-tailed correlational hypothesis (alternate) Write a two-tailed correlational hypothesis (alternate) Write a null hypothesis for a correlation Identify the two co-variables in a correlation Suggest how the co-variables can be operationalised/measured An alternative way of measuring the co-variables One strength and one weakness of the alternative measure Describe and evaluate an appropriate procedure Sketch an appropriate and labelled scattergraph Draw conclusions from a scattergraphObservation tick sheetStuff you must know for observations Identify two general strengths and two weaknesses of observations Describe what is meant by participant observation Identify a strength and weakness of participant observation Describe what is meant by structured observation Identify a strength and weakness of structured observation Describe what is meant by time sampling Identify a strength and weakness of time sampling Describe what is meant by time sampling Identify a strength and weakness of time sampling Describe and evaluate an appropriate procedure Know what is meant by reliability Know how to ensure that an observation has inter-rater reliability Know what is meant by validity Know how to improve the validity of an observation Describe what is meant by quantitative data
Outline one strength and one weakness of collecting quantitative data using observational studies Describe what is meant by qualitative data Outline one strength and one weakness of collecting qualitative data using observational studies Ethical issues relating to observations and how these ethical issues can be overcome