Research Methods PSYA4 Science A branch of knowledge conducted on objective principles involving the systematised observation/experiment with phenomena.

. The major features of science Empiricism Information is gained through direct observation or experiment rather than by reasoned argument or unfounded beliefs. Objectivity scientists want to be objective in their observations and measurements i.e. their expectations should not affect what they record. Replicability one way to show the validity of any observation/ experiment is to repeat it. If the outcome is the same, the original results are confirmed, especially if the observations have been made by a different person. Control In order for the experiment to be a fair test all other conditions must be kept the same i.e. controlled. Theory Construction A theory is a collection of general principles that explain observations and facts.

The scientific process Induction OBSERVATIONS Deduction

TESTABLE HYPOTHESIS

CONDUCT A STUDY TO TEST THE HYPOTHESIS

DRAW CONCLUSIONS

refine

refine
PROPOSE THEORY

Induction involves reasoning from the particular to the general. In the inductive model, this leads scientists to develop hypotheses. Hypotheses are tested, possibly leading to new questions and new hypotheses. Eventually this data can be used to construct a theory.

Deduction involves reasoning from the general to the particular. The deductive model places theory construction at the beginning after making observations.

Can psychology claim to be a science? Scientific research is desirable Psychology is a science because it shares the goals of all sciences and uses the scientific method. Kuhns view Thomas Kuhn (1962) claimed that psychology could not be a science because it doesnt have a single paradigm (a shared set of assumptions) like other sciences. It has lots of paradigms e.g. cognitive, physiological, behaviourist, evolutionary etc. Kuhn suggested it was a pre-science. Lack of objectivity and control. Objectivity is a key goal of science but in psychology, the object of study reacts to the researcher and this leads to problems such as experimenter bias and demand characteristics.

Are the goals of science appropriate for psychology? Nomothetic vs idiographic Science takes the nomothetic approach, looking to make generalisations about people and find similarities. Idiographic (treating each case individually, cant make generalisations) Psychological approaches to treating mental illness have had modest success, suggesting that the goals of science arent always appropriate. Qualititative research - triangulation Some psychologists use more qualitative methods of conducting research. These can be triangulated (findings from these different methods are compared with each other as a means of verifying them and making them objective (objectivity is a key goal of science).

Scientific approach is reductionist and deterministic Reductionist reduces complex phenomena to simple ones making it difficult to pick out patterns or make conclusions Deterministic searches for casual relationships and so oversimplifies the relationship between causes and effects.

The role of peer review

Peer review is the assessment of scientific work by others who are experts in the same field. The intention of peer reviewing is to ensure that any research conducted and published is of high quality. Peer review has 3 main purposes: Allocation of research funding Publication in scientific journals Research Assessment Exercise

Research published on the Internet requires new solutions to maintain the quality of information. Disadvantage of peer review 1. Unachievable ideal It isnt always possible to find an appropriate expert to review a research proposal or report. This means that poor research might be passed because the reviewer didnt really understand it. 2. Anonymity is usually used so that reviewers are honest and objective. But can have the opposite effect if they use anonymity to hide rival research. 3. Publication bias peer review favours the publication of positive results, possibly because editors want research that has important implications to increase the position of their journal against competitors. 4. Preserving the status quo- Peer review results in a preference for research that goes with existing theory rather than dissenting or unconventional work. Q. Discuss the process of peer review Answer: A peer review is when an expert in the field being written about reviews the article to judge its quality. This is usually unpaid and is done anonymously to encourage objectivity and honesty, though at the same time, this may have the opposite effect some reviewers might use it as an opportunity to prevent competing researchers from publishing work. Peer reviews may be an ideal, whereas in practice there are lots of problems, for example it slows publication down and may prevent unusual, new work being published. Some people doubt whether peer review can really prevent the publication of fraudulent research. The advent of the internet means that a lot of research and academic comment is being published without official peer reviews than before, though systems are evolving on the internet where everyone really has a chance to offer their opinions and police the quality of research.

Conventions for reporting psychological investigations

Scientific journals contain research reports which tend to be organised into the following sections: 1) Abstract A summary of the study covering the aims//hypothesis, method/procedures, results and conclusions. 2) Introduction/Aim What the researchers intent to investigate. This often includes a review of previous research (theories and studies), explaining why the researchers intend to conduct this particular study. The researchers may state their research predictions/ a hypothesis or hypotheses. 3) Method A description of what the researchers did, providing enough information for replication of the study. This section includes information about the participants (the sample), the testing environment, the procedures used to collect data, and any instructions given to participants before (the brief) and afterwards (the debrief). 4) Results This section contains what the researchers found, often called statistical data, which includes descriptive statistics (tables, averages and graphs) and inferential statistics (the use of statistical tests to determine how significant the results are). 5) Discussion The researchers offer explanations of the behaviours they observed and might also consider the implications of the results and make suggestions for future research. 6) References The full details of any journal articles or books that are mentioned.

Reliability & Validity Reliability a measure of consistency both within a set of scores or items (internal reliability) and also over time, such that it is possible to obtain the same results on later occasions when the measure is used (external reliability). The reliability of an experiment can be determined through replication (copying the experiment). Inter-observer reliability The extent to which there is agreement between two or more observers involved in observations of a behaviour. A general rule is that if the total number of agreements divided by the total number of observations is more than 80%, than the data has high inter-observer reliability. Validity refers to how true something is as an explanation of behaviour. It involves issues of control, realism and generalisability. Control refers to the extent to which any variable is held constant or regulated by a researcher. If an experimenter fails to control extraneous variables then the results of the study will be meaningless. The experimenter may claim that the IV caused a change in the DV but this may in fact be caused by something else - the extraneous variable(s). Consequently, the experimenter may not have actually tested what he/she intended to test. Instead the influence of a different variable has been tested. Mundane realism refers to how a study mirrors the real world. The simulated task environment is realistic to the degree to which experiences encountered in the environment will occur in the real world. Internal validity Whether the study has tested what it set out to test; the degree to which the observed effect was due to the experimental manipulation rather than other factors such as extraneous variables. External validity The degree to which a research finding can be generalised: to other settings (ecological validity); to other groups of people (population validity); different times (historical validity).

Ethical issue Informed consent Participants have the right to be given comprehensive, information concerning the nature and purpose of the research and their role in it, in order that they can make an informed decision about whether to participate.

Ethical Issues How to deal with it Participants are asked to formally indicate their agreement to participate and this should be based on comprehensive information concerning the nature and purpose of the research and their role in it. An alternative is to gain presumptive consent. Researchers can also offer the right to withdraw.

Limitations If a participant is given information concerning the nature and purpose of a study this may invalidate the purpose of the study. Even if researchers have sought and obtained informed consent, that does not guarantee that participants really understand what they have let themselves in for. The problem with the presumptive consent is that what people expect that they will or will not mind can be different from actually experiencing it. Cost-benefit decisions are flawed because they involve subjective judgements, and the costs are not always apparent until after the study. Debriefing cant turn the clock back a participant may still feel embarrassed or have lowered self-esteem. Participants may feel they shouldnt withdraw because it will spoil the study. In many studies participants are paid or rewarded in some way e.g. university students may be given course credits so they may not feel able to withdraw

Deception Where a participant is not told the true aims of the study (e.g. what participation will involve) and thus cannot give truly informed consent. The right to withdraw Participants should have the right to withdraw from participating in a study if they are uncomfortable in any way, and should also have the right to refuse permission for the researcher to use any data they produced. Protection from harm During a research study, participants should not experience negative physical or psychological effects, such as physical injury, lowered self-esteem or embarrassment. Confidentiality A participants right to have personal information protected. Privacy A persons right to control the flow of information about themselves. Contrast with confidentiality.

The need for deception should be approved by an ethics committee, weighing up benefits (of the study) against costs (to participants). Participants should be fully debriefed after the study and offered the opportunity to withhold their data. Participants should be informed at the beginning of a study that they have the right to withdraw.

Avoid any risks greater than everyday life. Stop the study.

Researchers are not always able to accurately predict the risks of taking part in a study.

Researchers should not record the names of any participants; they should use numbers or false names.

Do not observe anyone without their informed consent unless it is in a public place. Participants may be asked to give their retrospective consent or withhold their data.

It is sometimes possible to work out who the participants were on the basis of the information provided e.g. the location of a school. Therefore, in practice, confidentiality may not be possible. There is no universal agreement about what constitutes a public place. Not everyone may feel this is acceptable e.g. lovers on a park bench.

Ethical consideration in psychological research Ethics committee A group of people within a research institution that must approve a study before it begins. Ethical guidelines Concrete, quasi-legal documents that help to guide conduct within psychology by establishing principles for standard practice and competence. Debriefing After the research has taken place to inform the participants of the true nature of the study. Participants should be offered the opportunity to discuss any concerns they may have and to withdraw their data from the study as if they never took part. Presumptive consent A method of dealing with a lack of informed consent or deception, by asking a group of people who are similar to the participants whether they would agree to take part in a study. If this group of people consents to the procedures in the proposed study, it is presumed that the real participants would also have agreed. Code of conduct The current BPS (British Psychological Organisation) identifies 4 ethical principles and advice on how these should be dealt with: 1. Respect for the dignity and worth of all people. This includes standards of privacy, confidentiality and informed consent. Observations of behaviour in public without informed consent are only acceptable in situations where the people being studied would reasonably expect to be observed by strangers. Intentional deception (lack of informed consent) is only acceptable when it is necessary to protect the integrity of research and when the nature of the deception is disclosed to participants at the earliest opportunity. One way to judge deception is to consider whether participants are likely to object or show unease when debriefed, in which case the deception may be judged as unacceptable. Participants should be aware of the right to withdraw from the research at any time. 2. Competence Psychologists should maintain high standards in their professional work. 3. Responsibility Psychologists have a responsibility to their clients, to the general public and to the science of Psychology. This includes protecting participants from physical and psychological harm as well as debriefing at the conclusion of their participation to inform clients of the nature and conclusions of the research, to identify any unpredicted harm and arrange for assistance if needed. 4. Integrity Psychologists should be honest and accurate. This includes reporting the findings of any research accurately and acknowledging any potential limitations. It also includes bringing instances of misconduct by other psychologists to the attention of the BPS.

Ethical consideration in psychological research Ethical issues with animals Reasons for carrying out research on animals: Because they are fascinating to study in their own right and the research might benefit animals Animals offer the opportunity for greater control and objectivity in research procedures. Might use animals in research where you cant use humans. Animals have been exposed to various procedures and events that would not be possible with humans. Human beings and animals have a common evolutionary past to justify conclusions drawn from experiments involving one, to the other. However, it can be argued that animals tested under stressful conditions may provide little useful information. Moral justification Sentient beings do animals experience pain and emotions? There is evidence that animals respond to pain but this might not be the same as conscious awareness. However, there is evidence that some animals other than primates have self-awareness. Speciesism Singer (1990) argues that discrimination on the basis of species is no different from racial or gender discrimination and thus suggests that the use of animals is an example of speciesism. However Gray (1991) says we have a duty to take care of humans and so speciesism isnt the equivalent to e.g. racism. Animal rights Singer claims that if animal research can alleviate pain and suffering it is justifiable. However, Regan (1984) says there are no circumstances under which research is acceptable because animals have a right to be treated with respect. However this can be challenged by saying animals have no responsibilities in societies i.e. as citizens and therefore have no rights. Existing constraints Animal research is strictly controlled. The BPS publishes guidelines for research with animals. The Animals Act (1986) requires that animal research only takes place at licensed laboratories with licensed researchers on licensed projects. Licenses are only given if: Potential results are enough to justify the use of animals The research cannot be done using non-animal methods The minimum number of animals will be used Any discomfort or suffering is kept to a minimum by appropriate use of anaesthetic or pain killers The 3Rs were proposed by Russell and Birch (1959): Reduction use fewer animals Replacement where possible use alternative methods such as brain scans Refinement use improved techniques to reduce stress.

Research methods and concepts Experiments All experiments involve and IV (independent variable) and DV (dependent variable). The IV is varied in order to see how this affects the DV, thus demonstrating a casual relationship. As far as possible, all other variables are controlled, so any changes in the DV are due to the IV rather than extraneous variable. Laboratory experiment an experiment carried out in a controlled setting. Lab experiments tend to have high internal validity and low external validity, though this isnt always the case. Field experiment A controlled experiment conducted outside a laboratory. The IV is still manipulated by the experimenter, and therefore casual relationships can be demonstrated. Field experiments tend to have lower internal validity (more difficult to control EVs) and higher external validity (greater mundane realism). Participants are usually unaware that they are participating in an experiment, thus reducing participant effects. Natural experiment A research method in which the experimenter cannot manipulate the independent variable directly, but where it varies naturally and the effect can be observed on a dependent variable. Strictly speaking, an experiment involves the deliberate manipulation of an IV by the experimenter, so casual conclusions cannot be drawn from a natural experiment. In addition, participants are not randomly allocated to conditions in a natural experiment, which may reduce validity. Comparing lab, field and natural experiments Lab experiment When is it used? To investigate casual relationships under controlled conditions Advantages Well controlled; extraneous variables are minimised, thus higher internal validity. Can be easily replicated (repeated) to check if the same results occur which supports the external validity of the results. Less artificial, usually higher mundane realism and thus higher internal validity. Avoids participant effects (because participants not aware of study), which may increase internal validity. Allows research where IV cant be manipulated for ethical practice or practical reasons, e.g. studies of privation. Enables psychologists to study realproblems such as the effects of disaster on health (increased mundane realism). Disadvantages Artificial contrived situation where participants may not behave as they do in everyday life because of a lack of mundane realism, participant effects, investigator effects and demand characteristics. This reduces internal validity.

Field experiment When is it used? To investigate casual relationships in more natural surroundings

Natural experiment When is it used? To investigate casual relationships in situations where IV cannot be manipulated by an experimenter

Extraneous variable less easy to control because the experiment is taking place in the real world, thus reducing internal validity. There may still be demand characteristics, for example the way an IV is operationalised may convey the experimental hypothesis to participants. Cannot demonstrate casual relationships because IV not directly manipulated Inevitably many extraneous variables (e.g. lack of random allocation), which are a threat to validity. Can only be used where conditions vary naturally. Participants may be aware of being studied causing participant effects, investigator effects and demand characteristics.

Experimental design Experimental design A set of procedures used to control the influence of factors such as participant variables in an experiment Repeated measure design Each participant takes part in every condition under test Repeated measure design Weaknesses In an experiment, one condition may be more difficult than the other. E.g. if you wanted to see whether people remember more in the morning or afternoon, you could give participants a memory test in the morning and a different test in the afternoon. However, it might be that the participants did better in the morning because the test was easier than the test in the afternoon. In this case, changes in the DV would be due to an extraneous variable (easier test) rather than the IV. Participants may guess the purpose of the experiment, which may affect their behaviour. Ways of dealing with the weaknesses You can make sure the tests are equivalent. Create a list of 40 words and randomly allocate them to two lists so that both lists are equivalent.

You can use a cover up story about the purpose of the test to try to prevent them guessing what it is about (SINGLE BLIND)

E.g. some participants might do worse purposely Single blind A type of research design in which on the second test because they want it to the participant is not aware of the research aims appear as if they work less well in the afternoon. or of which condition of the experiment they are receiving. The order of the conditions may affect performance (order effect) Order effect In a repeated measures design an extraneous variable arising from the order in which conditions are presented, e.g. practice effect or fatigue effect. E.g. participants may do better on the second test because of a practice effect or because they are less anxious or they may do worse on the second test because of being bored with doing the same test again (boredom or fatigue effect). Independent groups design Participants are allocated to two (or more) groups representing different experimental conditions. Allocation is usually done using random techniques. Independent groups design Weaknesses No control of participant variables (i.e. the different abilities or characteristics of each participant) You need twice as many participants. Ways of dealing with the weaknesses Randomly allocate participants to conditions which distributes participant variables evenly. Be prepared to spend more time and money. Use counterbalancing Counterbalancing An experimental technique used to overcome order effects. Counterbalancing ensures that each condition is tested first or second in equal amounts.

Experimental design Matched pairs design Pairs of participants are matched in terms of key variables such as age and IQ. One member of each pair is places in the experimental group and the other member in the control group. Matched pairs design Weaknesses Time consuming to match participants on key variables. May not be able to control all participant variables because you can only match on variables known to be relevant, but it could be that others are more important. E.g. in a memory experiment, you might match on memory abilities but later find that some participants had been involved in a teaching programme to boost memory skills and you should have matched on this. Ways to deal with the weaknesses Restrict matching variables to make it easier. Conduct a pilot study to consider key variables. Pilot study A small-scale trial run of a study to test any aspects of the design, with a view to making improvements.

Self-report techniques Self-report methods are non-experimental (because the person is reporting their own beliefs/thoughts), and include questionnaires and interviews. Questionnaire Data are collected through the use of written questions. Interview A research method or technique that involves a face-to-face interaction with another individual and results in the collection of data. Structured interview Any interview in which the questions are decided in advance. Unstructured interview The interview starts out with some general aims and possibly some questions, and lets the interviewees answers guide the subsequent questions. Open questions Questions that invite respondents to provide their own answers rather than select one of those provided. Tend to produce qualititative data. Closed questions Questions that have a range of answers from which respondents select one e.g. yes/no. Produces quantitative data. Answers are easier to analyse then those for open questions. Questionnaires Respondents record their own answers Advantages Can be easily repeated so that data can be collected from large numbers of people relatively cheap and quickly. Respondents may feel more willing to reveal personal/confidential information than in an interview. Disadvantages Answers may not be truthful, for example because of leading questions (indicating direction of answer to give) and social desirability bias(presenting yourself in a better light). The sample may be biased because only certain kinds of people fill in questionnaires literate individuals who are also willing to spend time filling in a questionnaire and returning it. Can be easily repeated because the The interviewers expectations may questions are standardised. influence the answers the interviewee gives (interviewee bias) Requires less interviewing skill than Interviewee bias The effect of an interviewees an unstructured interview. expectations, communicated unconsciously (not on More easy to analyse than an purpose), on a respondents behaviour. unstructured interview because Reliability may be affected by the same answers are more predictable. interviewer behaving differently on different occasions or different interviewers asking different questions (low inter-interviewer reliability). More detailed information can More affected by interviewer bias than generally be obtained from each structured interviews because in an respondent than in a structured unstructured interview the interviewer is interview. developing new questions on the spot which might be less objective. Can access information that may not be revealed by predetermined Requires well-trained interviewers, which questions. makes it more expensive to produce reliable interviews compared with structured interviewers which dont require specialist interviewers.

Structured interview Questions are predetermined

Unstructured interview Interviewer develops questions in response to respondents answers

Self-report techniques Validity External validity The external validity of questionnaires and interviews concerns the extent to which the findings can be generalised to other situations and other people. Another factor is the representativeness of the sample used to collect data. Internal validity The internal validity of self-report techniques is related to whether the questionnaire or interview really measures what it intended to measure. The most common ways to assess this are: Face validity: does the test look as if it is measuring what the researcher intended to measure? Concurrent validity: This can be established by comparing the current questionnaire or test with a previously established test on the same topic. Participants take both tests and then the two test scores are compared. Improving validity Validity is improved by assessing the validity of a questionnaire or interview. If validity is low then: External validity: Use a more appropriate sampling method to improve population validity because then the findings can be generalised to a wider population. Internal validity: If one or more measures of internal validity are low, then the items on the questionnaire/interview/test need to be revised in order to produce a better match between scores on the new test and an established one. Reliability Internal reliability is a measure of the extent to which something is consistent within itself. E.g. all questions on an IQ test should be measuring the same thing. External reliability is a measure of consistency over several different occasions. E.g. if an interviewer conducted an interview, and then conducted the same interview with the same person a week later, the outcome should be the same otherwise the interview is not reliable. Inter-interviewer reliability the extent to which two interviewers produce the same outcome from an interview. Assessing reliability Internal reliability Split-half method: A single group of participants all take a test once. Their answers to the test questions are divided in half. This is done by, for e.g., comparing all answers to odd number answers with all answers to even numbers. The individuals scores on both halves of the test should be very similar. The two scores can be compared by calculating a correlation coefficient. Test-retest method: A group of participants is given a test or questionnaire or interview once and then again sometime later (when they have had the chance to forget it). The answers can be compared and should be the same. If the test produces scores, these can be compared by calculating a correlation coefficient.

External reliability

Self-report techniques

Improving reliability Internal reliability: Remove those items which are most inconsistent. The only way to do this is by trial and error remove one test item and see if the split-half correlation coefficient improves. If it does, then the removed item should be permanently left out. Ethical issues Deception about true research aims may sometimes be necessary to collect truthful data. Psychological harm respondents may feel distressed by certain questions or having to think about some sensitive topics. Privacy questions may be related to sensitive and personal issues, invading an individuals privacy. Confidentiality must be respected; names and personal details should not be revealed without permission. No personal data may be stored.

Observational studies Different kinds of observation Naturalistic observation A research method carried out in a naturalistic setting, in which the investigator does not interfere in any way but merely observes the behaviour(s) in question, though this is likely to involve the use of structured observations. Controlled observation A form of investigation in which behaviour is observed but under controlled conditions, in contrast with a naturalistic observation. Observational techniques Observation techniques The application of systematic methods of observation in an observational study, experiment, or other study Observational research aims to be objective and rigorous. For this reason it is preferable to use observational techniques.

Structured observations Structured observations An observer uses various systems to organise observations, such as behavioural categories and sampling procedures. The researcher uses various systems to organise observations Behavioural categories How to record the behaviour you are interested in. Sampling procedures Who you are observing and when.

Unstructured observations Unstructured observations An observer records all relevant behaviour but has no system. This technique may be chosen because the behaviour to be studied is largely unpredictable.

Behavioural categories Dividing a target behaviour into a subset of behaviours. This can be done using a behaviour checklist or a coding system. Sampling procedures Sampling The process of taking a sample which is intended to be a representative selection of a target population. Continuous observation is not possible because there would be too much data to record, therefore there must be a systematic method of sampling observations: Event sampling An observational technique in which a count is kept of the number of times a certain behaviour (event) occurs. Time sampling An observational technique in which the observer records behaviour in a given time frame, e.g. noting what a target individual is doing every 30 seconds. You may select one or more categories from a checklist.

Designing observational research In order to make systematic and objective observations, researchers need to develop behavioural categories. The method is called a coding system or behaviour checklist. Coding system A systematic method for recording observations in which individual behaviours are given a code for ease of recording. Behaviour checklist A list of the behaviours to be recorded during an observational study. Covert observations Observing people without their knowledge, e.g. using one-way mirrors. Knowing that behaviour is observed is likely to alter a participants behaviour.

Distinctions Methods and technique All research involves making observations. In some research the overall method is observational where the emphasis is on observing a relatively unconstrained segment of a persons freely chosen behaviour. However observational techniques are used in almost all studies, even experiments. Controlled and naturalistic Both kinds of observation use systematic methods to record observations i.e. there is control over how the observations are made. Control over the environment is only true in a controlled observation. Participant and non-participant In many cases the observer is just watching the behaviour of others and acts as a non-participant. In some studies, observers also participate which may affect their objectivity. Overt and covert One-way mirrors are used to prevent participants being aware that they are being observed. This is called covert or undisclosed observation. Knowing that your behaviour is being observed is likely to alter your behaviour, and therefore observers often try to be as unnoticeable (low-profile) as possible, though this has ethical implications. Naturalistic observation and natural experiment Both involve naturally occurring variables that have not been manipulated by the researcher. However, in a natural experiment, there is an IV and its effect on a DV is observed so that we can draw tentative casual conclusions. In a naturalistic observation there is no IV.

Evaluating observational research Validity External validity Observational studies are likely to be high in ecological validity because they involve more natural behaviours. Population validity may be a problem e.g. children are only observed in middle class homes because we cannot generalise such findings to children from all classes. Internal validity Observations will not be valid (or reliable) if the coding system/behaviour checklist is flawed. For example, some observations may belong in more than one category or some behaviour may not be codeable. The validity of observations is affected by observer bias. Observer bias In observational studies, there is the danger that what someone observes is influenced by their expectations. This reduces the validity of the observations. Improving validity A researcher can improve validity by conducting observations in varied settings with varied participants, which makes the findings more generalisable to other settings and other people. The researcher can also use more than one observer to reduce observer bias and averaging data across observers. Reliability Reliability A measure of consistency both within a set of scores or items (internal reliability) and also over time, such that it is possible to obtain the same results on subsequent occasions when the measure is used (external reliability). The reliability of an experiment can be determined through replication. Reliability of observations Inter-observer reliability The extent to which the agreement between two or more observers involved in observations of a behaviour. Observations should be consistent, which means that ideally two observers should produce the same record. The extent to which two or more observers agree is called inter-observer reliability. A general rule is that if there is more than 80% agreement on the observations, the data have interobserver reliability. Total agreements > 80% Total observations Improving reliability Observers should be trained in the use of a coding system/behaviour checklist. They should practice using it and discuss their observations. The investigator can then check the reliability of their observations.

Evaluating observational research Ethical issues 1) Informed consent participants are observed without their knowledge. 2) Invasion of privacy/confidentiality 3) Deception use of one way mirrors When participants are aware that they are being observed, informed consent, right to withdraw etc still applies. Dealing with ethical issues Ethics committees can be used to approve observational designs, and researchers should consult ethical guidelines. One BPS guideline concerning observational research is that this type of research is acceptable when those observed would be expected to be observed by strangers. Researchers should be aware that it is not acceptable to invade privacy of people who, even in a normally public place, may believe they are unobserved. Evaluation of observational research Advantages What people say they do is often different from what they actually do, so observations may be more valid than questionnaires/interviews. Gives a more realistic picture of spontaneous behaviour. It is likely to have high ecological validity. Provides a means of conducting preliminary investigation in a new area of research, to produce hypotheses for future investigations. Disadvantages There can be little or no control of extraneous variables which may mean that something unknown to the observer may account for the behaviour observed. Observer bias what the observer expects may influence what they see. Low inter-observer reliability different observers see different things. If participants dont know they are being observed there are ethical problems such as deception and invasion of privacy. If participants do know they are being observed they may alter their behaviour.

Correlational analysis Correlation/correlation analysis Determining the extent of a relationship between two variables Positive correlation Both variables increase together Zero correlation Variables may not be linked at all Negative correlation As one variable increases, the other decreases. Linear correlation Where all the values lie in a straight line from bottom left to the top right. Curvilinear correlation The relationship between the variables is not linear but curved Scattergram A graphical representation of the relationship (i.e. the correlation) between two sets of scores. Significance A statistical term indicating that the research findings are sufficiently strong for us to accept the research hypothesis under test. Intervening variable A variable that comes between two other variables which is used to explain the relationship between those two variables. For e.g. if a positive relationship is found between ice cream sales and violence this may be explained by an intervening variable such as heat, which causes the increase in ice cream sales and the increase in violence. Evaluation of studies using correlational analysis Advantages Can be used when it would be unethical or impractical to manipulate variables and can make use of existing data If correlation is significant then further investigation is justified If correlation is not significant then you can probably rule out a casual relationship As with experiments, the procedures can be reported again which means that the findings can be confirmed Disadvantages People often misinterpret correlations and assume that a cause and effect have been found, whereas this is not possible. There may be other, unknown variables (intervening variables) that can explain why the co-variables being studied are linked. As with experiments, may lack internal/external validity, for e.g. method used to measure IQ may lack validity or sample used may lack generalisability

Case studies Case study A research method that involves a detailed study of a single individual, institution or event. They are usually longitudinal (follow an individual over a period of time) Advantages The method offers rich, in-depth data so information that may be overlooked using other methods is likely to be identified. Can be used to investigate instances of human behaviour and experiences that are rare, for example, investigating cases of children locked in a room throughout childhood to see what effects such deprivation has on emotional development. It would not be ethical to generate such conditions experimentally. The complex interaction of many factors can be studied, in contrast with experiments where many variables are held constant. Disadvantages It is difficult to generalise from individual cases as each one has unique characteristics. It is often necessary to use recollection of past events as part of the case history and such evidence may be unreliable. Researchers may lack objectivity as they get to know the case, or because theoretical bias may lead them to overlook aspects of the findings. There are important ethical issues such as confidentiality many cases are easily identifiable because of their unique characteristics, even when real names are not given.

Content analysis Content analysis A kind of observational study in which behaviour is observed indirectly in written or verbal material such as interviews, conversations, books, diaries or TV programmes. Behaviour is categorised (qualitative analyses) and may be counted (quantitative analysis). Advantages Has high ecological validity because they are based on direct observations of what people actually do; real communications which are current and relevant, such as recent newspapers or childrens books in print. When sources can be retained or accessed by others (e.g. back copies of magazines or videos of people giving speeches) findings can be replicated and so tested for reliability. Disadvantages Observer bias reduces the objectivity and validity of findings because different observers may interpret the meaning of the behavioural categories differently.

Likely to be culture-bias because interpretation of verbal or written content will be affected by the language and culture of the observer and behavioural categories used.

Other research methods & techniques Attrition The loss of participants from a study over time which is likely to leave a biased sample or a sample which is too small Cohort effects One group of participants (cohort) may have unique characteristics because of timespecefic experiences during the development of its members, such as children born just before the war who experienced poor diets in infancy due to rationing. This can affect both cross-sectional studies (because one group is not comparable with another) or longitudinal studies (because the group studied is not typical)

Control condition In a repeated measures experiment, the condition that provides a baseline measure of behaviour without the experimental treatment (IV), so that the effect of the experimental treatment may be assessed. Control group In an independent groups experiment, a group of participants who receive no treatment. Their behaviour acts as a baseline against which the effect of the IV may be measured. Cross-cultural study A kind of natural experiment in which the IV is different cultural practices and the DV is a behaviour such as attachment. This enables researchers to investigate the effects of nature and nurture. + The technique enables psychologists to see whether some behaviours are universal i.e. not affected by cultural differences. Researchers may use tests or procedures that have been developed in the US that are not equally valid in their country. This may make the individuals in the researchers culture appear abnormal(imposed etic) The group of participants may not be representative of the culture, and yet researchers might make generalisations about the whole culture/ country. Cross-sectional study One group of participants of a young age are compared with another, older group of participants , with a view to finding out the influence of age on the behaviour in question. Two groups of participants might be quite different. The participant variables in a cross-sectional design are not controlled, because the participants in each group are different. Cohort effects. In a cross-sectional study, cohort effects might produce spurious results. E.g. one cross-sectional study might compare the IQs of people aged 20+ with 80+ and find that the IQs of the second group were much lower, concluding that ageing led to a decrease in IQ. The real reason however, might be because the 80+ had lower IQs when they were in their 20s because of e.g. poorer diet. Effect size A measure of the strength of the relationship between the two variables. Experimental condition In a repeated measures design, the condition containing the experimental treatment (IV). Experimental group In an independent groups design, a group of participants who receives the experimental treatment (the IV). Imposed etic A technique or theory is developed in one culture and then used to study the behaviour of people in a different culture which has different norms, values, experiences etc. Longitudinal study Observations of the same items over a long period of time. Such studies usually aim to compare the same individuals at different ages, in which case the IV is age. A longitudinal study might also observe a school or other institutions over a long period of time. Attrition. The people who drop out are likely to have particular characteristics (e.g. be the ones who are less motivated or more unhappy), which leaves a biased sample. People are likely to become aware of the research aims and their behaviour may be affected. Cohort effects. You might not be able to generalise the findings from a study that looks at only one cohort because of the unique characteristics of that cohort.

Other research methods & techniques Meta-analysis A researcher looks at the findings from a number of different studies in order to reach a general conclusion about a particular hypothesis. + Analysing the results from a group of studies rather than from just one study can allow more reliable conclusions to be drawn. Often studies produce contradictory results (e.g. some studies might find no effect, some studies a small effect, and others a larger effect). A meta-analysis will allow us to reach an overall conclusion. The research designs in the different studies sampled may vary considerably, which means that the studies are not truly comparable, and therefore the conclusions are not always valid. Role play A controlled observation in which participants are asked to imagine how they would behave in certain situations, and act out the part. This method has the advantage of allowing the study of certain behaviours that might be unethical or difficult to find in the real world. + Enables researchers to study behaviour that might otherwise be impractical or unethical to observe. E.g Zimbardo prison and guard experiment. People might not act as they do in real life. E.g. in zimbardos study, the participants acting as guard might have been following what they thought was guard-like behaviour from films. But if they acted like real-life guards, they may have acted more in accordance with personal principles rather than according to social norms.

Sampling Systematic sample A method of obtaining a representative sample by selecting e.g. every fifth or tenth person. This can be a random sample if the first person is selected using a random method; you then select every fifth or tenth person after that. Target population The group of people that the researcher is interested in. The group of people from whom a sample is drawn. The group of people about whom generalisations can be made. Volunteer bias A form of sampling bias because volunteer participants are usually more highly motivated than randomly selected participants. Advantages The easiest method because you can just use the first participant you can find, which means it takes less time to locate your sample than if using one of the other techniques. Access to a variety of participants (e.g. all the people who read a newspaper) which would make the sample more representative and less biased. Unbiased. All members of the target population have an equal chance of selection. Disadvantages It is biased because the sample is drawn from a small part of the target population.

Opportunity sample Participants are selected by using those people who are most easily available at the time of the study.

Volunteer sample A sample of participants produced by a sampling technique that replies solely on volunteers to make up the sample. Also called a self-selected sample.

Sample is biased because participants are likely to be more highly motivated / have more time on their hands (volunteer bias).

Random sample A sample of participants produced by using a random technique such that every member of the target population being tested has an equal chance of being selected. Stratified and quota samples Groups of participants are selected according to their frequency in the population in order to obtain a representative sample. Groups or strata are selected from the target population that need to be represented, and then individuals are selected form the strata. If this final sample is carried out using a random technique it is a stratified sample, otherwise it is a quota sample. Snowball sampling Staring off with one or two people and asking them to direct you to some other people with the same characteristic involved e.g. two people with eating disorders directing you to more people they know with eating disorders.

The researcher might end up with a biased sample (e.g. more boys than girls) because the sample is too small.

More representative than an opportunity sample because there is equal representation of sub-groups.

Although the sample represents sub-groups, each quota taken may be biased in other ways, e.g. if you use opportunity sampling you only have access to certain sections of the target population.

This is useful when conducting research with participants who are not easy to identify.

Sample is prone to bias because researchers may only contact people within a limited section of the population.

Quantitative data Analysis

Quantitative data analysis Any means of representing trends from numerical data, such as measures of central tendency Quantitative Data that represent how much or how long, or how many etc there are of something i.e. a behaviour is measured in numbers or quantities. Measures of central tendency Measures of central tendency A descriptive statistic that provides information about a typical response for a data set. Mean Mean The arithmetic average of a group of scores. Takes the values of all the data into account The mean is calculated by adding up all the numbers and dividing by the number of numbers. + It makes of use of the values of all the data It can be misrepresentative of the data as a whole if there are extreme values It cannot be used with nominal data Median Median The middle value in a set of scores when the scores are placed in rank order + Not affected by extreme values - Not as sensitive as the mean because not all values are reflected in the median. Mode Mode The value that is most common +Useful when the data are in categories i.e. nominal data - Not a useful way of describing data when there are several modes. Types of data Nominal The data are in separate categories, such as grouping people according to e.g. their favourite football team, gender etc. Ordinal Data is ordered in some way e.g. asking people to put a list of football teams in order of their liking. The difference between each item is not the same, i.e. the individual may like the first item a lot more than the second, but there might only be a small difference between the items ranked second and third.(ranking e.g. grades) Interval Data are measured using units of equal intervals, such as when counting correct answers or using any public unit of measurement e.g. temperature Ratio There is a true zero point as in most measures of physical quantities.

Measures of dispersion Measures of dispersion A descriptive statistic that provides information about how spread-out a set of scores is. Range & Standard deviation Range The difference between the highest and lowest score in a data set Standard deviation shows the amount of variation in a data set. It assesses the spread of data around the mean. Range Advantages Provides you with direct information Easy to calculate Disadvantages Affected by extreme values Doesnt take into account the number of observations in the data set May hide some of the characteristics of the data set (e.g. influence of extreme values)

Standard deviation

More precise measure of dispersion because all values are taken into account

Graphical representation Graphs Tables The numbers you collect are called raw data. Bar chart The height of the bar represents frequency (how often something happens). Shows data in categories but is suitable for numbers. Line graph The y axis represents frequency but x values must be continuous i.e. numerical data but not in categories e.g. favourite football teams. Scatter gram Suitable for correlational data which can show positive correlation (pattern going from bottom left to top right) , negative correlative(pattern going from top left to bottom right) and zero correlation (no pattern)

Qualitative Data Qualitative Data that expresses the quality of things descriptions, words, meanings, pictures, texts etc. Qualitative data cannot be counted or quantified but they can be turned into quantitative data by placing them into categories. Methods of evaluating qualitative data analysis Coding Coding is the process of identifying categories, themes, phrases or keywords that may be found in any set of data. The categories/ themes etc are decided in either of the following ways: 1.Top-down approach (thematic analysis) codes represent ideas and concepts from an existing theory/explanation. 2. Bottom-up approach (grounded theory) where the codes/categories emerge from the data. Summarising the data The behavioural categories can be used later to summarise the data. It is also possible to count frequency of occurrences in each category, thus qualitative data is turned into quantitative data. Finally a researcher can draw conclusions. Validity and reflexivity Validity is a measure of the extent to which reality has been achieved. The qualitative approach denies the existence of any one real world, i.e. each individuals subjective view is reality(each individual thinks of reality in different ways) Qualitative researchers acknowledge the need for reflexivity the recognition that a researchers attitudes, biases etc have an unavoidable influence on the research they are conducting. The impact of reflexivity cannot be avoided but it can be monitored and reported. The validity of qualitative research findings can be demonstrated using triangulation, comparing the results from a variety of different studies of the same thing or person. The studies are likely to have used different methodologies. If the results agree, this supports their validity. If the results are different, this can lead to further research to enhance our understanding. Reliability is a component of validity and can be checked, for e.g. by looking at the inter-rater reliability when more than one person has coded the data. Qualitative Data VS Quantitative Data Advantages Quantitative Easier to analyse because the data are given in Data numbers that can be summarised using measures of central tendency and dispersion as well as simple graphs Can produce neat conclusions because numerical data reduces the variety of possibilities. Qualitative Represents the true complexities of human behaviour Data Gains access to thoughts and feelings that may not be assessed using quantitative methods with closed questions Provides rich details of how people behave because participants are given a free range to express themselves.

Disadvantages Oversimplifies reality and human experience (statistically significant but humanly insignificant)

More difficult to detect patterns and draw conclusions because of the large amount of data usually collected Subjective analysis can be affected by personal expectations and beliefs (though quantitative methods may only appear to be objective but are equally affected by bias)

Inferential analysis, probability and significance Operationalisation providing variables in a form that can be easily tested. Researchers start by identifying what they intend to study (the aims) and then make a formal statement of their expectations using a hypothesis. A hypothesis can be directional or nondirectional. EXAMPLE OF OPERATIONALISTAION: Hypothesis: People remember more when they study in short bursts than when studying for longer sessions. Step1: Operationalise the IV and Dv What is meant by remember more , short bursts and longer sessions? In order to test this hypothesis we need to be specific about behaviour or operations that can be measured or manipulated for the IV and DV. Operationalising the IV Short bursts of study can be operationalised as 10-minute study sessions repeated three times over a period of three hours. Longer sessions can be operationalised as one 30-minute session. Operationalising the DV Remember more can be operationalised by deciding how to assess recall we could do this by giving a chapter of a book to revise and giving them a set of questions to see how many questions they get right. Final fully operationalised hypothesis: People get more questions right on a test of recall when they study in short bursts (ten minutes at a time repeated three times) than when studying for longer sessions (one 30-minute session). Directional (one tailed) & non-directional hypothesis (two tailed) Directional hypothesis states the direction of the predicted difference between two conditions or two groups of participants. E.g. People who do homework without the TV on, produce better results than those who do homework with the TV on. Directional hypotheses are also called one-tailed because they are only concerned with one possibility. Non-directional hypothesis predicts simply that there will be a difference between two conditions or two groups of participants, without stating the direction of the difference. E.g. People who do homework with the TV on, produce different results than those who do homework with no TV on. The non-directional hypothesis covers both possibilities and this is two tailed.

Type I and Type II errors A type I error is in situations where the null hypothesis is rejected when in fact, the results are due to chance and should have been accepted. A type II error is said to have occurred when researchers fail to find any significant effect when in fact one does exist i.e. the null hypothesis is accepted when it should have been rejected. Type II errors can be the result of a poor research design, faulty sampling or random error. Null hypothesis is accepted Type I error Correct Null hypothesis is rejected Correct Type II error

Reject null hypothesis Accept null hypothesis

The likelihood of making a Type I and type II error depends on the significance level set in the research. If we set a more stringent significance level than the standard 5% level(0.05), such as 1%(0.01), then we are more likely to make a Type II error. On the other hand, if we accepted a 10% (0.1) significance level then we are more likely to make a Type I error.

Inferential analysis, probability and significance Probability Inferential statistics allows psychologists to draw conclusions based on the probability that a particular pattern of results could have arisen by chance. However, if it could not have arisen by chance, then the pattern is described as significant (the extent to which something is unusual). Probability is used to decide whether or not a result is significant. Significance Psychologists use the term significance to refer to results that are very unlikely to have occurred by chance alone. Chance In order to work out whether a difference is significant you use inferential tests. They let you work out, for a given probability, if a pattern in data from a study could have happened by chance or whether the effect occurred because there is a real difference/correlation in the populations from which the samples were drawn. The chosen value of p is called the significance level. In general, psychologists use a probability of p<0.05, which means that there is less than or equal to a 5% possibility that the results did occur by chance i.e. at least a 5% probability that the results occurred even if there was no real difference/association between the populations from which the samples were drawn. Using inferential statistics Different tests: Spearmans rho Chi-square Mann-Whitney Wilcoxon T Factors affecting choice of test: What research design is used The level of measurement used (nominal, ordinal, interval, ratio) If there is a correlation

Observed and critical values Test statistic a single number produced by doing arithmetic calculations from the data collected in a study by each inferential test e.g. test statistic for spearmans test is called rho, where-as for Mann-Whitneys test it is U. Observed value The rho value calculated for any set of data because it is based on the observations made. Critical value The value a test statistic must reach in order for the null hypothesis to be rejected. To find the appropriate critical value in a table you need to know: 1. Degrees of freedom (df) In most cases you get this value by looking at the number of participants in the study (N). In the case of the chi-square test you calculate df on the basis of how many cells there are. 2. One-tailed or two-tailed test If the hypothesis was a directional hypothesis, then you use a one-tailed test. If it was non-directional you use a two-tailed test. 3. Significance level usually p<0.05 The importance of R Some tests are significant when the observed value is equal to or exceeds the critical value, for others it is the reverse. One way is to remember is to see if there is a letter R in the name of the test. If there is an R then the observed value should be gReateR than the critical value (e.g. for Spearmans and Chi-square) for significance to be shown. If there is no R (e.g. Mann-Whitney and Wilcoxon) then the observed value should be less than the critical value for significance to be shown.

Inferential tests: Spearmans rho Spearmans rho is used to determine whether the correlation between two co-variables is significant or not.

When to use spearmans rank correlation (rho) test The hypothesis predicts a correlation between two co-variables The two sets of data are pairs of scores from one person or thing = i.e. they are related. The data are ordinal or interval (i.e. not nominal)

Stating the conclusion (Remember Spearmans Rho has a letter R in it so) Observed value for the test must be equal to or greater than the critical value for significance to be shown. If significance is shown (i.e. observer value greater than or equal to critical value) then reject the null hypothesis and accept alternative hypothesis. If the observed value is less than the critical value then you accept the null hypothesis because significance is not shown. EXAMPLE Alternative hypothesis: The digit ratio between the index finger and ring finger is positively correlated to numeracy skills. (This is a directional hypothesis, therefore requiring a one-tailed test). Null hypothesis: There is no correlation between digit ratio and numeracy skills. Conclusion - The observed value (0.58) is greater than the critical value (0.564) we could reject the null hypothesis (at p<0.05) and therefore conclude that digit ratio is correlated with numeracy (because it is significant).

Chi-Square (X) Test This test deals with nominal data. We use this test when we have counted how many occurrences there are in each category called frequency data. When to use the Chi-Square (X) Test The hypothesis predicts a difference between two conditions or an association between covariables. The sets of data must be independent (no individual should have a score in more than one cell). The data are in frequencies (i.e. nominal). Frequencies must not be percentages. NOTE: This test is unreliable when the expected frequencies (i.e. the ones you calculate) fall below 5 in any cell because you need atleast 20 participants for a 2 x 2 contingency table)

Stating the conclusion (Remember Chi Square has a letter R in it so) Observed value for the test must be equal to or greater than the critical value for significance to be shown. If significance is shown (i.e. observer value greater than or equal to critical value) then reject the null hypothesis and accept alternative hypothesis. If observed value is less than the critical value, accept the null hypothesis. EXAMPLE Alternative hypothesis: Certain parental styles are associated with higher self-esteem in adolescence. (This is a non-directional hypothesis and therefore requires a two-tailed test). Null hypothesis: There is no association between parental style and self-esteem in adolescence.

Conclusion - As the observed value (1.984) is less than the critical value (3.84) we must accept the null hypothesis (at p<0.05) and therefore conclude that there is no association between parental style and self-esteem in adolescence.

Mann-Whitney U Test

When to use the Mann-Whitney U Test The hypothesis predicts a difference between two sets of data The two sets of data are from separate groups of participants = independent groups The data are interval or ordinal (i.e. not nominal).

Stating the conclusion The name Mann-Whitney does not have the letter R in it so the observed value of U must be equal to or less than the critical value in order to be significant. If the observed value is less than the critical value( i.e. significant), we must reject the null hypothesis and accept the alternative hypothesis. If the observed value is greater than the critical value (i.e. not significant) we must accept the null hypothesis and reject the alternative hypothesis EXAMPLE Alternative hypothesis: Male participants interviewed on a high bridge give higher ratings of the attractiveness of a female interviewer than those interviewed on a low bridge (This is a directional hypothesis therefore requires a one-tailed test) Null hypothesis: There is no difference in the ratings of attractiveness given by those interviewed on a high or low bridge. Conclusion - As the observed value (16.5) is less than the critical value (41), we can reject the null hypothesis (at p<0.05) and therefore conclude that participants interviewed on a high bridge give higher ratings of attractiveness to a female interviewer than those interviewed on a low bridge.

Wilcoxon T Test

This test is appropriate for tests of difference where pairs of data are related, such as when repeated measures design has been used. When to use the Wilcoxon T Test The hypothesis predicts a difference between two sets of data. The two sets of data are pairs of scores form one person (or a matched pair) = related. The data are ordinal or interval (i.e. not nominal) Stating the conclusion (Remember the name Wilcoxon does not have the letter R in it so) the observed value of T must be equal to or less than the critical value for significance to be shown. If the observed value if greater than the critical value (i.e. not significant) we must accept the null hypothesis and reject the alternative hypothesis. If the observed value is less than the critical value (i.e. is significant) we must reject the null hypothesis and accept the alternative hypothesis. EXAMPLE Alternative hypothesis: Participants rate the more frequently seen face as more likeable than the less frequently seen face (This is a directional hypothesis and therefore requires a one-tailed test). Null hypothesis: There is no difference in the likeability score for faces seen more or less often. Conclusion As the observed value (18.5) is greater than the critical value (13) we must accept the null hypothesis (at p<0.05) and conclude that there is no difference in the likeability score for faces seen more or less often.

Example scenarios for what test to use?

Read the following scenarios. For each one, decide: a whether the researchers are looking for a difference between performance in two conditions or a relationship/association between two variables b what level of measurement is being used c if they are looking for a difference, whether they have used a repeated or independent measures design. On the basis of your answers to these questions, decide which statistical test would be appropriate to analyse the data.

1.
An A-level Psychology teacher is interested in whether there is a relationship between the amount of time students spend in paid employment and how well they do at A-Level. She asks students how many hours they spend each week at work, and translates their predicted A-level grades into a points score (A=10, B=8etc.).

2.
Does the nature of the material to be learned affect how well it is remembered? Participants are asked to learn a list of 30 words, half of which are related to emotions (e.g. love, hate) and half of which are neutral words, matched for length and approximate frequency in the English language. They are asked to recall as many words as they can. Their scores for the two kinds of words are compared.

3.
A primary school teacher is interested in the language skills of her pupils, and in particular whether there is a link between size of vocabulary and spelling ability. She gives them a vocabulary test, which requires them to match a series of words with their meanings, and compares their scores with those from a spelling test.

4.
Researchers are interested in the effect on helping behaviour of watching altruistic behaviour on TV. One group of children is shown a video where friendly cartoon animals help old ladies across the road and help to carry their shopping, while another group is shown a Tom and Jerry cartoon. All the children are then exposed to situations where they could offer help. Observers decide whether each child is very helpful, quite helpful or not helpful, and compare the ratings for the two groups.

5.
In a memory experiment, participants are asked to learn pairs of words. One group is asked to learn the words by repeating them over and over again, while another group is asked to create a visual image to link the words. They are then tested by being shown the first word of each pair, and asked to supply the second word. The number of words correctly remembered in each condition is compared.

6.
Researchers are interested in whether information about party policy affects people's voting intentions. A sample of voters in Anytown is asked whether or not they are likely to vote for the Monster Raving Loony Party. They then attend a lecture on Monster Raving Loony Party policy, following which they are again asked whether they are likely to vote for this party. Researchers note how many people who had not originally intended to vote for the party changed their minds after attending the lecture.

7.
Do children change their opinion of school subjects when they have spent some time at a particular school? Children entering Dotheboys Hall are asked to rate how interesting they find English, Maths and Science, using a scale of 1 (extremely boring) to 10 (absolutely great). They are asked to rate these subjects again when they have been at the school for a year, to assess whether there has been any change.

8.
Gumrot plc have produced a new line in sweets called Dinosaur Droppings. They want to know whether there is a difference between the two possible packaging designs they have developed in terms of their appeal to girls and to boys. Sweets are displayed in both packaging designs. The numbers of each kind bought by girls and by boys are recorded and compared.

9.
Researchers are interested in the truth of well-known sayings, in particular 'elephants never forget' and 'a memory like a sieve'. Participants are hypnotised and given a post-hypnotic suggestion either that they are an elephant or that they are a sieve. They are then asked to read and memorise a story. Recall for each group is compared, using a questionnaire on the content of the storv.

10.

ANSWERS
Psychologists at Oxford Clinic for the Seriously Bewildered are interested in whether their therapy methods, which involve teaching clients A-level Psychology, can be shown to be effective. Staff are asked to rate a group of clients on the amount of bewildered behaviour they show, both before and after studying A-level Psychology.
1.Spearmans Rho 2.Wlicoxon signed ranks 3.Spearmans Rho 4.Chi squared 5.Mann Whitney U 6.Sign test 7.Wlicoxon signed ranks 8.Chi squared 9. Mann Whitney U 10. Wlicoxon signed ranks

Summary of which test to use

OR When deciding which test is appropriate in any situation, you can ask yourself three questions in the diagram below:

Nominal data?

CHI-SQUARE (X) TEST

Correlation ?

SPEARMANS RHO TEST

Independent groups?

MANN-WHITNEY TEST

WILCOXON T TEST