Psychology: research methods (booklet 1)unit 2

Validity

The ability of a test to measure what it is intended to measure

Population validity (External validity)

extent where the sample is a representative of a wider or different population.

ecological validity (external validity) /mundane realism

The extent to which a study is realistic or representative of real life.

Historical validity (external validity)

Study is representative of different points in history.

double-blind procedure (overcoming issues of validity)

When neither researcher nor participant know the true aims of the research. Prevents researcher bias, that way the researcher doesn't unknowingly give away the aim or any hints to the participant. This also reduces the chances of demand characteristics (where participant finds out aim and changes behaviour accordingly to what they they best fit in the best light)

Single-blind procedure (overcoming issues of validity)

Involves the researcher knowing the true aim however the participant not knowing the research hypothesis until after the role is complete.

anonymity/confidentiality

Anonymity:

Identity of participants not known to researchers)

Confidentiality:

Only the research investigators know the identity of the participants

Reduces impact of social desirability bias.

Internal validity & external validity

Internal:

The study measures what it intents to measure as the confounding variables have been controlled, therefore will not effect the DV of any participant, meaning the DV is only caused by the IV which is how it is supposed to be.

external:

External validity is the extent to which you can generalise the findings of a study to other situations, people, settings, and measures. The aim of scientific research is to produce generalisable knowledge about the real world. Without high external validity, you cannot apply results from the laboratory to other people or the real world.

Reliability

consistency of measurement: Research findings is produced consistently over a number of investigations.

inter-rater reliability (type of reliability)

Extent to which 2 or more observers agree

-producing the same data

-if 80% or over of both of the data of the observers are the same then it is reliable.

-to ensure that you can use a coding system or a checklist etc.

Internal reliability (internal consistency) (type of reliability)

a study participant gives a consistent pattern of answers, no matter how the researcher has phrased the question.

-to ensure you can use a "split half method" which is where you have 1 group and half the number of participants now you have 2 subgroups who will both take a test, both subgroups even and odd answers are compared to see if they correlate, if they do it is reliable.

External reliability (type of reliability)

A measure of consistency over several different occasions.

- to test this you ca use the "test- re-test method" where a group of participants are given a test/interview/questionnaire and after sometime has passed they are re-tested to see if their answers are similar or the same as before.

-Problem with this is that someone may get better at the test or whatever so the answers may vary so it could be inconsistent.

Accuracy of measuring tool (issues of reliability)

If you have a bad measuring tool your internal and external validity will be less reliable.

Operationalisation of variables

Making sure you said how measured and operationalised your variables specifically otherwise the person can copy or replicate it properly.

Order/practice effect (issues of reliability)

The more you do something the better you get at it. Same with whatever conditions you are in if you are constantly placed under the same conditions then it will slowly with time have less of a effect on you.

Demand characteristics (issues of reliability)

A clue which makes the participants unconsciously aware of the aims of the study,which will then consequently alter natural behaviour.

Standardisation

Keeping everything the same including the conditions (IV) for all of the participants so that the investigation is fair.

Random sampling (types of sampling)

Participants are chosen from the sampling frame randomly that way everyone has an equal chance of being chosen.

-ignore repeats

-not bias

-however may not be a representation of everyone.

Snowball sample (types of sampling)

Ask one participant and they tell their friend and recruit them.

-helpful when investigating a rare type of characteristic or behaviour.

-initial participant may not know enough people.

-takes more time to complete.

Self-selected sample (types of sampling)

Participants volunteer and self select themselves.

-no ethical issues and no screw you effect.

-not enough participants.

-may be unable to form appropriate subgroup within target population.

Stratified sampling (types of sampling)

dividing a population into a series of subgroups and taking random samples from within each group.

-All of my subgroups will be representatives.

- it is more harder for researchers to administer in comparison

-selected person may not be a representative of sub-group

quota sampling (types of sampling)

Where the target population is divided into groups and at convenience you would choose who you want.

-more biased

-shorter time compared to stratified sampling

Systematic sampling (types of sampling)

Every nth item in the target population is selected.

-not biased

-researcher may not be biased however system could be flawed.

Order effects (evaluating the experimental design)

Order effects occur when participating in one condition may affect how a participant performs in another. Participants performance in the experimental condition may be better, not because of the IV being manipulated but because of practice effects after completing the control condition. On the other hand, their performance in the experimental condition may be worse, again not because of the IV being manipulated but because of fatigue or the boredom.

Demand characteristics (evaluating the experimental design)

Demand characteristics are a type of confounding variable where participants unconsciously work out the aim and act differently (either through social desirability or the screw you effect). If a participant takes part in more than one condition, they are more likely to be able to work out the aim of the experiment.

The number of participants (evaluating the experimental design)

The number of participants required to collect a sufficient amount of data depends on what sort of experimental design you use. Using fewer participants is a practical choice in research and can save a researcher a great deal of time and effort.

Repeated measures (experimental design)

There is one group of participants. The participants take part in both conditions of the experiment. And repeat it.

Advantages:

-straightforward

-number of participants

-participants variables

Disadvantages:

- order of effects may influence how the subject performs based based of the timing

-demand characteristics

Method of dealing with disadvantages:

-counterbalancing: - participant split into two groups one group completes control condition first then the other group completes the experimental. They then switch.

Independent groups (experimental design)

Participants only take part in one condition e.g. 30 participants in total 15 only take part in experimental whilst the other 15 take part in the controlled condition.

Advantages:

-demand characteristics less likely only because they are only taking part in one.

-order of effect.

Disadvantages:

- number of participant you need different people

-participant variables sometimes not comparable.

Method for handling disadvantage:

-randomly allocating participants to conditions e.g. names in a hat.

matched pairs design (experimental design)

Experiential and control condition are deliberately similar e.g. monozygotic twins e.g each twin only take part in one condition.

Advantages

-no order of effects

-avoids participant variables

Disadvantages:

-costly

-not practical

-requires a big sample of similar groups of people

Method of dealing with disadvantage:

-Use pilot study: Helps consider the key variable that might be important to matching

You could also restrict number of variables to make it easier.

The location of research

The location that a psychologist chooses to conduct their research is one which can depend on the sort of method they are using, or it might rely on the approach or area of psychology that they are studying.

• There are three main locations of research that you need to know:

o Laboratory experiments

o Field experiments

o Online experiments

Difference studies (Quasi-experiments)

In a difference study the apparent IV is also naturally occurring and the DV may be measured in a laboratory.

The key feature is that the IV has not been made to vary by anyone.

It is simply a difference between people that exists.

So it isn't actually a variable at all. Such 'difference studies' are included as quasi-experiments.

independant variable

variable that is changed in an experiment

dependent variable

the variable that may change in response to manipulations of the independent variable. It is also the thing that you measure

External validity

the extent to which the results of a study can be generalized to other situations and to other people

Lack of standardisation of procedures (issues of reliability)

Making sure you are doing the same things each time so your results are reliable and follow a certain protocol.

Researcher bias (issues of reliability)

Unconsciously having a preference which will effect the validity of the researcher findings.

Social desirbility bias (issues of reliability)

The participants change their behaviour in order to be perceived in a better light.

Opportunity sampling (types of sampling)

Participants are chosen because they are available so by convenience.

-it is easy and quick

-you end up with a biased sample

-could end up with a similar group of people

main features when evaluating the experimental design:

-Order effects

- Demand characteristics

-The number of participants

•-Participantvariables

Participant variables (evaluating the experimental design)

Participant variables are the individual characteristics that all participants have. Participant variables such as age, gender, personality, intelligence or education background could affect how we respond in a piece of research.

Operationalised hypothesis

Defining precisely how you intend to measure the DV and alter the conditions of the IV

Extraneous variable (EV)

Variable that is not bing manipulated

by the researcher but affect the (Dv) results of each and ALL

participants equally. The variable i usually inconsistent

and link to a design flaw. Researcher avoid this

as they are undesirable , hard to control and

a effort should be made so minimise these.

Confounding Variables (CV)

The same as extraneous variables except they effect the DV of only SOME of the participants

Hypothesis

A testable prediction, often implied by a theory

Alternative Hypothesis (Ha)

This predicts what will happen if our theories on human behaviour is true.

Alternative Hypothesis: directional/one-tailed

States the direction its going in. E.g. girls do better than boys in maths.

Alternative Hypothesis: non-directional/two-tailed

Does not clearly state the direction it is going in.

Null Hypothesis (H0)

Assumes there is no significant relationship difference between the variables being studied.

Laboratory part 1 (location of research)

Conducting research in a laboratory is the most scientific way to conduct research.

• A laboratory is a room equipped to allow scientific research and measurement.

• Mostpsychologicalresearchthatgetsconducted in a laboratory tends to be experimental in nature, but many researchers also conduct observational research in a laboratory setting.

laboratory part 2 (location of research)

Laboratory: an environment which can be controlled by the researcher. In particular a researcher wishes to control extraneous variables and, in an experiment, needs to manipulate the independent variables. This is easy in a controlled environment.

Advantages:

-order of effects

-you can replicate it

- internal validity

Disadvantages:

-External validity

-mundane realism

E.g. loftus and palmer study 1974

Field experiments part 1 (location of research)

experiments conducted in natural settings rather than in the laboratory

A hospital might seem like quite a controlled environment, making it difficult to distinguish between a laboratory and the field - one way to do this is to think of whether the participants have gone to the research (laboratory) or the researcher has gone to the participants (field).

Field experiments 2 (location of research)

E.g. Piliavin et al. (1969) conducted some very famous field experiments on 'helping behaviour' in the New York Subway.

Advantages:

-Mundane realism

- ecological validity

-useful application

Disadvantages:

-Ethics

-less access to tools

Online (location of research)

Online: refers to being connected via internet to another source. You may be connected to a website or could be using an app on a mobile phone. Data can be collected from individuals.

The method of conducting research has only arisen in the last couple of decades.

• Being online allows researchers to be able to access participants via the internet or on social networking tools. This kind of research is often questionnaires but can be experimental.

Online part 2 (location of research)

Advantages:

-Lots of participants so larger sample

-easy convenient

Disadvantages:

-methods are limited

-less tools

-similar demographic

Quasi-experiments

A quasi-experiment is a not 'true' experiment because the the independent variable has not been deliberately manipulated.

Meaning the changes of the dependant variable are not caused by the independent variable.

Natural experiment (Quasi-experiments)

A natural experiment is conducted when it is not possible, for ethical or practical reasons, to deliberately manipulate an IV.

The DV may be tested in a laboratory.

Natural experiments (Quasi-experiments)

Example 1: Effects of watching violence - Berkowitz (1970)

• compiled monthly FBI crime statistics on the frequency of

violent crimes from January 1960 to December 1966.

• He reported a steep increase in the number of violent crimes being recorded following November 1963.

• This, he proposed, was the result of the American population being exposed to television footage of John F. Kennedy's assassination.

• In this study the IV was exposure to television footage of Kennedy's assassination.

Natural experiments (Quasi-experiments)

Example 2: Effects of TV - Charlton et al. (2000)

• Before 1995 people living on the small island of St Helena in

the middle of the Atlantic had no TV.

• The arrival of TV gave researchers a chance to see how exposure to Western programmes might influence behaviour

• Overall Charlton et al. (2000) found no difference in either pro- or anti-social behaviour after the introduction of Western TV.

• In this study the IV was no TV and later exposure to TV.

Difference studies (Quasi-experiments)

Example 1: Gender differences - Sheridan and King (1972)

• tested obedience by asking male participants to give genuine

electric shocks of increasing advantage to a puppy.

• They found that 54% of male participants delivered the maximum (non-fatal) shock, but the obedience rate for females was a staggering 100%!

• The IV in this study was gender - a difference that cannot be manipulated and thus not a 'true' IV.

Example 2: Olfactory abilities of non-sighted individuals - Rosenbluth et al. (2000)

• Compared the olfactory (smell) abilities of sighted and non- sighted children.

• One group of 30 non-sighted children were matched to 30 sighted children in terms of their age, sex and ethnicity.

• The researchers found that although the two groups of children did not differ significantly in their olfactory abilities, the non-sighted children scored higher (12.1 out of a maximum of 25) compared to the sighted children (10.4) when labelling smells.

participant observation

a naturalistic observation in which the observer becomes a participant in the group being observed

Non-participant observation

when the researcher observes behavior without participating in that behavior

Strengths and weaknesses of participant observations

+Provides real insight into the participants being studied

- Subjective (researcher is part of the group being studied - open to biased interpretations of behaviour from the researcher)

Strengths and weaknesses of non-participant observations

+More objective (not part of the group being studied - so less chance of biased interpretations of behaviour)

-may lose valuable insight into the participants behaviour (e.g. they won't be aware of the context of the p's behaviour)

Unstructured observations (structuring observations)

Recording all behaviour with no system of doing so. Often used in pilot studies to identify behaviour categories for the actual study.

Structured observations (structuring observations)

the investigator sets up a laboratory situation that evokes the behavior of interest so that every participant has an equal opportunity to display the response

Behavioural categories

When a target behaviour is broken up into components that are observable and measurable

sampling procedure

method of choosing the people to be studied

observational studies

Researchers carefully and systematically observe and record behavior without interfering with behavior.

Event sampling

the observer records all instances of a particular behavior during a specified time period

Time sampling

a measurement of the presence or absence of behavior within specific time intervals

Questionnaires (self report technique)

A research method that includes a list of written questions, which generate closed and/or open answers. These can be used in person, online or through other methods, e.g. postal.

• Questionnairescanproducequantitativeor qualitative data or a mixture of both.

Structured interview (self-report technique)

A research method that includes standardised questions (like in a questionnaire), which are known as an interview schedule, and are usually asked face to face.

• The interviewer has a pre-prepared set of questions that are asked in a fixed order.

• Pre-determined questions are used that elicit a verbal response.

• An example of a structured interview is Herjanic's (1977) research, which showed that behaviour and school performance were greater indicators of psychiatric disturbance than neurotic and somatic symptoms.

Semi-structured interview (self-report technique)

A research method which involves asking participants questions, usually face to face. These can be in the form of an interview schedule, however they could also include follow up questions to expand on answers of the questions asked.

• This is sometimes called a clinical interview because it can be likened to the kind of interview you might have with a doctor.

• They start with some predetermined questions, but further questions are developed as a response to your answers.

• An example of a semi-structured interview is Aira et al (2003), who found that when doctors discussed alcohol consumption as an 'addiction', reported levels from patients were lower than when alcohol consumption was discussed as a 'lifestyle factor'.

Questionnaires strengths

-They can be easily distributed to a large number of people which means data can be collected quickly and conveniently. This increases time and cost efficiency as well as increasing the sample size and thus representativeness of the target population.

- Participants may feel happier and less embarrassed disclosing confidential information if the researcher is not physically present and their responses are kept anonymous.

Questionnaires weaknesses

1. People may produce 'response set' answers by ticking the same answer box to all questions.

2. The sample may be biased because they can only be completed by literate people with time to fill them in.

Self-report techniques

An non-experimental technique or method used to find out about peoples behaviours

Structured interview (weaknesses)

- can't ask for elaboration

Semi-structured interview weaknesses

-limited space to explore themes that weren't planned on beforehand

-one-on-one situation can appear artificial and lower ecological validity

-data analysis/collection is very time consuming

Semi structured interview strengths

lots of detail generated, fairly flexible and sensitive, reliable, easy to analyze, stratified

Structured interview strengths

- High Validity and Reliability

- You can compare responses

- Can identify if there is or isnt a probem

Ethical issue

an identifiable problem, situation, or opportunity that requires a person to choose from among several actions that may be evaluated as right or wrong, ethical or unethical

Risk to participant values,relationships, status, or privacy

Requires psychologists to not just consider search as participant a part of a research study.

Working with vulnerable individuals

If researchers want to investigate and perform studies on vulnerable groups who may not completely understand what is being asked of them, they need to ensure that they have gained fully informed consent from the individual who is legally responsible. Children under 16 need to have their parent's permission to take part in a study. Adults with mental illnesses or learning

difficulties likewise need to have consent from their guardian.

Working with animals

Research with non-human species is strictly controlled. Restrictions on type of animal, care, number of animals required etc. are all controlled by ethical guidelines and laws such as the Animals Act (1986).

Valid consent (researcher point of view)

Valid consent means revealing the true aims of the study - or at least telling participants what is actually going to happen. However, revealing the details might cause participants to guess the aims of a study. For example, a psychologist might want to investigate whether people obey a male teacher more than a female teacher. If the participants are told the aim of this experiment before the study takes place, it might change the way they behave - they might try to be equally obedient to both. Researchers therefore may not always want to reveal the true aims or even the full details of what is going to happen.

Valid consent (participant point view)

Participants should be told what they will be required to do in the study so that they can make an informed decision about whether they wish to participate. This is a basic human right, established during the Nuremberg war trials. In the Second World War Nazi doctors conducted various experiments on prisoners without their consent and the war trials afterwards decided that consent should be a right for participants involved in any study.

Even if researchers have obtained valid consent, that does not guarantee that participants really do understand what they have let themselves in for. Epstein and Lasagna (1969) found that only a third of participants volunteering for an experiment really understood what they had agreed to take part in.

Another problem is the requirement for the researcher to point out any likely benefits or risks of participation. Researchers are not always able to accurately predict the risks or benefits of taking part in a study.

Deception (researcher point of view)

It can be necessary to deceive participants about the true aims of a study otherwise participants might alter their behaviour and the study could be meaningless. A distinction, however, should be made between withholding some of the details of the research aims (reasonably acceptable) and deliberately providing false information (less acceptable).

Deception (participant point of view)

Deception is unethical - the researcher should not deceive anyone without good cause. Perhaps more importantly, deception prevents participants being able to give valid consent. They may agree to participate without really knowing what they have let themselves in for and they might be quite distressed by the experience.

Deception can also lead people to see psychologists as untrustworthy. It might further mean that a participant may not want to take part in psychological research in the future.

Risk of harm (researcher point of view)

Studying some of the more important questions in psychology may involve a degree of risk of harm (psychological or physical) to participants. It is also difficult to predict the outcome of certain procedures (such as in the Milgram obedience study; see page 162) - therefore it is difficult to guarantee protection from any risk of harm.

Risk of harm (participants point of view)

Nothing should happen to them during a study that causes harm. There are many ways harm can be caused to participants, some physical (e.g. getting them to smoke, or drink coffee excessively) and some psychological (e.g. making them feel inadequate, or embarrassing them). It is considered acceptable if the risk of harm is no greater than a participant would be likely to experience in ordinary life, and if participants are in the same state after a study as they were

before, unless they have given their valid consent to be treated otherwise.

Confidentiality (researcher point of view)

It may be difficult to protect confidentiality because the researcher wishes to publish the findings. A researcher may guarantee anonymity (withholding participants' names), but even then it may be obvious who has been involved in a study. For example, knowing that a study of children in hospital was conducted on the Isle of Wight could permit some people to be able to identify participants because the target group has been narrowed down.

Confidentiality (participant point of view)

The Data Protection Act makes confidentiality a legal right. It is only acceptable for personal data to be recorded if the data are not made available in a form that identifies the participants.

Privacy (researcher point of view)

It may be difficult to avoid invasion of privacy when studying participants without their awareness, for example in a field experiment

Privacy (participant point of view)

People do not expect to be observed by others in certain situations, for example when in the privacy of their own homes, while they might expect this when sitting on a park bench in public.

Valid consent (dealing with specific issues)

the issue of valid consent is dealt with by asking participants to give their valid consent; the issue of risk of harm is dealt with by ensuring that people are in the same state after a study as they were before, unless they have consented.

• The BPS code of practice identifies issues and at the same time suggests how these issues are dealt with.

• The most obvious way of dealing with ethical issues is through the use of codes of practice (guidelines) produced by a professional organisation. Psychologists, like other scientists, have other ways of dealing with ethical issues.

Right to withdraw (dealing with ethical issues)

If a participant begins to feel uncomfortable or distressed they should be given the right to withdraw.

• This is especially important if a participant has been deceived about the aims and/or procedures.

• However, even if a participant has been fully informed, the actual experience of taking part may turn out to be rather different, so they should be able to withdraw.

Depriefing (dealing with ethical issues)

At the end of a study participants are given various kinds of information about the study they took part in.

• The aim of such debriefing is to inform the participants of the true nature of the study and to restore them to the same state they were in at the start of the study.

• Debriefing is not an ethical issue; it is a means of dealing with ethical issues such as deception and psychological harm and any unforeseen psychological harm that may have arisen in the research.

Depriefing part 2 (dealing with ethical issues )

If participants were deceived about the true aims of the study, they are told what the aims were during the debriefing.

• Participants may not have been deceived but might not have been told the full details of the study, for example there may have been several conditions in the study, such as a placebo condition.

• During the debriefing they should be given any other information about the study so they understand their role more fully. If the participants are harmed in any way (e.g. stressed or made to feel awkward or embarrassed), they should be offered reassurance that their behaviour was normal and, where necessary, be offered extra counselling.

• Participants should be offered the right to withhold their data if they object to having participated.

Ethics committees (dealing with ethical issues)

Most institutions where research takes place have an ethics

committee which must approve any study before it begins.

• The committee looks at all possible ethical issues raised in any research proposal and at how the researcher suggests that the issues will be dealt with, weighing up the benefits of the research against the possible costs to the participants.

• Members of the committee often include lay people as well as experts in the field.

Punishment (dealing with ethical issues)

If a psychologist does behave in an unethical manner, such as conducting unacceptable research, then the BPS reviews the research and may decide to ban the person from practising as a psychologist.

• It is not a legal matter (the psychologist won't be sent to prison).