scientific processes

research questions

broad questions about something a researcher wants to know

aims

• statement of the study’s purpose, takes one feature of the question to focus on and is broader, or less precise than a hypothesis

• eg ‘to see if age affects the duration of short term memory’

hypothesis

• theories tested by research, a testable and predictive statement often generated from a theory, statement of what researcher thinks it true

• either states a predicted difference between IV and DV (experimental hypothesis) or states a predicted relationship between variables (correlational analysis)

• should be fully operationalised, variables and how they’ll be measured must be clear and measurable in the hypothesis

• null hypothesis states there’s no effect in a study

one tailed vs two tailed hypotheses

• one tailed (directional)- states the direction of the predicted difference eg teenagers will sleep for more hours than adults

• two tailed (non directional)- predicts a difference between 2 conditions but doesn’t specify which direction the difference will be eg there will be a significant difference between…

population

a large group of people who are the focus of the researcher interest, from which a smaller sample is drawn eg all students at the uni of Sheffield

sample

• a group of people who take part in a research investigation, drawn from a target population and is presumed to represent it

• sampling techniques are used to select people from the population

• testing the whole population is expensive, time consuming and might not all be available so researcher selects a sample that should be representative of general population so findings can be generalised

bias

in context of sampling, when certain groups are over or under represented within the sample selected, limits the extent to which generalisations can be made to the target population

generalisation

the extent to which findings and conclusions from an investigation can be applied to the population, made possible is the if the sample of participants is representative of the population

sampling methods- opportunity sample

• representative samples are difficult to obtain, many researchers decide to select anyone who happens to be willing and able, the ones nearest or easiest to obtain

• researcher asks whoever is around at the time of study

• eg walking around school and selecting whoever you come across for a questionnaire

strength of opportunity sampling- quick method

• opportunity sampling is convenient as you just make use of the people who are closest

• makes it cheaper and one of the most popular sampling methods

limitation of opportunity sampling- inevitably biased

• the sample is unrepresentative of the target population as it is drawn from a very specific area eg one street in town

• means the findings cant be generalised

sampling methods- random sampling

• a sophisticated form of sampling where all members have an equal chance of being selected in a target population

• a complete list of all the members of a target population is obtained, all the members are assigned a number then the sample is generated in a random manner eg names in a hat, number generator

strength of random sampling- potentially unbiased

• researcher has no influence on who’s elected so cant choose participants who may support their hypothesis, means CVs/EVs are controlled

• enhances internal validity

limitation of random sampling- time consuming and may not work

• complete list of population is hard to get, sample may be unrepresentative

• also some participants may refuse to take part making it like a volunteer sample

sampling methods- volunteer sampling

• involves participants selecting themselves to be part of the sample

• eg a researcher may place an advert in a newspaper or notice board

strength of volunteer sampling- participants are willing

• participants have selected themselves and know how much time and effort is involved

• likely to engage more than people stopped in the street and is quicker and easier for researcher

limitation of volunteer sampling- volunteer bias

• participants may share certain traits eg want to be helpful

• respond to cues and generalisation limited

sampling methods- systematic sampling

• when every nth member of a target population is selected

• before participants are selected, a sampling frame is produced, this is a list of people in the target population organised in some way eg alphabetical

• a sampling system is chosen eg every 3rd person, nay be determined by a computer system to reduce bias

strength of systematic sampling- unbiased

• the first item is usually selected at random, once the selection system is chose the researcher has no influence over who’s chosen

• makes it objective

limitation of systematic sampling- time and effort

• complete list of population is required which is hard to get

• may as well use random sampling

sampling methods- stratified sampling

• the researcher first identifies the different strata (subgroups) that make up the population eg age or gender, and then works out the proportions needed for the sample to be representative

• finally, the participants that make up each stratum are selected using random sampling, the relative subgroups in the population are reflected in the sample

strength of stratified sampling- representative method

• the characteristics of the target population are represented, avoids researcher bias as once the population has been divided into strata the participants are randomly selected and cant be influenced by the researcher

• generalisability more likely than other methods

limitation of stratified sampling- stratification isn’t perfect

• strata cant reflect all the ways in which people are different

• complete representation isn’t possible

pilot study

• small scale, trial run version of an investigation that takes place before the real investigation, uses smaller numbers of participants

• designed to road test the procedure and allows researchers to spot any issues before carrying out the study

what can pilot studies be used for?

• can be used for a range of study types (experimental, self report, observation), particularly useful for questionnaires and interviews as researchers can try out questions and reword or remove confusing questions

• in observations it gives the researcher a chance to check their coding systems and train observers

• in the short run, allows researchers to identify any potential issues and modify the procedure, saves time and money in the long run

single blind procedures

• participants aren’t aware of the aim of the study or ‘kept blind’ to an aspect of the procedure eg if they receive a drug or a placebo

• used to reduce the effects of demand characteristics as participants won’t know which condition they’re a part of

• participants may report fake effects if they think they’ve had the drug, but this would be easy to spot if they actually had the placebo

double blind procedures

• neither the participants or researchers are aware of the aim of the study or an aspect of the procedure to reduce demand characteristics and investigator effects

• used in drug trials as treatment, drug or placebo, is administered by someone who doesn’t know which condition each participant is in

• also reduces the chance that researchers expectations will influence participant behaviour, researchers may be more likely to record observations that support the aim if they know the participant has had the drug

control groups and conditions

• in independent groups, group that receives the real drug is the experimental group, group receiving the placebo are the control group

• present for the purpose of comparison, acts as a baseline to help establish causation, if the change in behaviour of the experimental group is significantly greater than the control the researcher can conclude the cause of the effect was the IV

• control conditions used in repeated measures to have same purpose, each participant takes part twice, once in experimental condition, once in control condition

experimental design- independent groups

one group does condition A and second group does condition B, participants should be randomly allocated to groups

strength of independent groups- no order effects

• participants are only tested once so cant practice or become bored/tired

• this controls and important CV

strength of independent groups- will not guess aim

• participants are only tested once so are unlikely to guess research aims

• therefore behaviour may be more ‘natural’, higher realism

limitation of independent groups- participant variables

• the participants in the 2 groups are different, acting as an EV/CV

• may reduce the validity of the study

limitation of independent groups- less economical

• need twice as many participants as repeated measures for same data

• more time spent recruiting which is expensive

experimental design- repeated measures

same participants take part in all conditions of an experiment, the order of conditions should be counterbalanced to avoid order effects

strength of repeated measures- participant variables

• the person in both conditions has the same characteristics

• this controls an important CV

strength of repeated measures- fewer participants

• half the number of participants are needed than in independent groups

• less time spent recruiting participants

limitation of repeated measures- order effects are a problem

• participants may do better or worse when doing a similar task twice, also practice/fatigue effects

• reduces the validity of the results

limitation of repeated measures- participants may guess aims

• participants may change their behaviour, demand characteristics

• may reduce the validity of the results

experimental design- matched pairs

2 groups of participants are used but they’re also related to each other by being paired on participant variables that matter for the experiment

strength of matched pairs- participant variables

• participants matched on a variable that is relevant to the experiment

• this controls participant variables and enhances the validity of the results

strength of matched pairs- no order effects

• participants are only tested once so no practice or fatigue effects

• enhances the validity of the results

limitation of matched pairs- matching is not perfect

• matching is time consuming and cant control all relevant variables

• cant address all the relevant variables

limitation of matched pairs- more participants

• need twice as many participants as repeated measures for same data

• more time spent recruiting which is expensive

ways of recording data

1. unstructured observation- researcher records everything they see to produce detailed data on few participants, continuous recording

2. structured observation- identify target behaviours which are the main focus of the investigation using behavioural categories and sampling methods

observational design- behavioural categories

• the target behaviour to be observed should be broken up into a set of observable and measurable categories, similar to operationalism

• eg target behaviours of aggression in primates at the zoo could be shown in many different ways: hitting, throwing, shouting, barring teeth

strength of behavioural categories- improves inter-observer reliability

• before beginning the observation, the researcher needs to list all the possible ways a target behaviour may occur

• this list is given to each observer to record when they see a particular behaviour

limitation of behavioural categories- difficult to make clear and unambiguous

• categories should be self evident and not overlap, not always possible to achieve

• ‘smiling’ and ‘grinning’ would be poor categories

limitation of behavioural categories- dustbin categories

• all forms of behaviour should be in the list and not one ‘dustbin’

• ‘dumped’ behaviours go unrecorded

observational design- event sampling

• counting the number of times a particular behaviour, the event, occurs in a target individual or group

• eg counting the number of times someone puts their hand up in a lesson

strength of event sampling- useful for infrequent behaviour

• the researcher will still ‘pick up’ behaviours that don’t occur at regular intervals

• such behaviours could easily be missed using time sampling

limitation of event sampling- complex behaviours oversimplified

• if the event is too complex, important details may go unrecorded

• may affect the validity of the findings

observational design- time sampling

• recording behaviour within a pre established time frame, observations are made at regular intervals eg every 15 seconds

• eg focus on one student in the class and record what they are doing every 2 minutes

strength of time sampling- reduces the number of observations

• rather than recording everything that is seen, i.e continuous, data is recorded at certain intervals

• the observation is more structured and systematic

limitation of time sampling- may be unrepresentative

• the researcher may miss important details outside of the timescale

• may not reflect the whole behaviour

designing questionnaires- open questions

• questions where there is no fixed response, respondents are free to answer in any way they wish

• tends to produce qualitative data, respondent provides own answers expressed in words eg ‘why did you start smoking?’ would produce a range of personal answers

strength of open questions- responses aren’t restricted

• answers are more likely to provide, detailed, unexpected information

• likely to have more external validity than statistics

limitation of open questions- difficult to analyse and compare

• wider variety of answers than produced by closed questions

• may be forced to reduce data to statistics

designing questionnaires- closed questions

• questions where there is a fixed choice of responses determined by the question setter, often yes or no responses or a scale

• scales produce quantitative data eg ‘how many cigarettes do you smoke a day? 0-10, 11-20 etc’, yes/no eg ‘do you smoke?’ can be converted to quantitative

strength of closed questions- easier to analyse

• can produce graphs and charts for comparison

• makes it easier to draw conclusions

limitation of closed questions- responses are restricted

• participants are forced into an answer that may not represent true feelings

• reduces validity of the findings

types of scale

1. Likert scales- involved respondents indicating their agreement with a statement using a scale of usually 5 points from strongly agreed to strongly disagree

2. rating scales- ask respondents to identify a value that represents their strength of a feeling about a particular topic

3. fixed choice option- includes a list of possible options and respondents are required to indicate those that apply to them

considerations for writing good questionnaire questions

1. clarity- respondents shouldn’t be able to misinterpret any questions, could lead to poor quality data

2. avoid jargon- specialist, technical terms may be confusing to non specialists

3. avoid emotive language- can give away researchers views on the area of study

4. avoid leading questions- guides participants towards a particular answer, can introduce investigator bias

5. avoid double barrelled questions- contain 2 questions in 1, can cause issues if respondents only agree with half

6. avoid double negatives- can confuse respondents and encourage them to give an answer they don’t agree with

considerations for designing interviews

1. interview schedule- standardise list of questions that the interviewer needs to cover, can reduce interviewer bias, ideally interviews recorded and notes taken later

2. quiet room- encourage participant to open up in interview

3. rapport- neutral opening questions can relax the participant and establish rapport

4. ethics- remind interviewees answers will be treated in confidence, participants usually interviewed individually, group interviews can be useful in clinical settings

independent vs dependent variables

• independent variable- the variable the researcher manipulates

• dependent variable- the variable the researcher measures

• should be operationalised, made specific, measurable and observable in a research study, specifying exactly how it will be measured eg time taken to complete a puzzle in minutes

confounding variables

• a variable beside the IV that affects the DV

• one the researcher has find out has affected the research after the experiment has taken place

• if extraneous variables (extra variables to the ones tested) affect the results of the experiment they become confounding and spoil results

extraneous variables

• variable beside the IV which could affect the DV

• if they’re not controlled it goes on to affect the study and the results have been confounded (spoiled)

• the reason for this is that if all other variables have been controlled we can be sure that the IV is affecting the DV in the experiment

difference between confiunfong and extraneous variables

• the extraneous variables are extra or additional variables that get in the way before the experiment but is then controlled eg unequal mix of boys and girls is then divided equally

• the confounding variable is one the researcher found out affected the findings after the experiment eg eyesight, intelligence levels

control of research issues- random allocation

• participants are randomly allocated to groups in experiments

• an attempt to control for participant variables in independent groups that ensures each participant has the same chance of being in one condition as the other

• decreases systematic error, so individual differences in responses or ability are less likely to consistently affect results

control of research issues- counterbalancing

• an attempt to control for the effects of order in repeated measures, half the participants experience the conditions in one order and the other half in the opposite order

• used to deal with the extraneous effects caused by order effects

control of research issues- randomisation

• the use of chance when designing investigations to control for the effect of bias eg allocating participants to conditions randomly

• keeps the research as objective as possible

control of research issues- standardisation

• using exactly the same formalised procedures for all participants in a research study, otherwise differences become EV’s

• means no participant receives an unfair advantage or is treated any differently to others, allows research to be replicated increasing it's reliability

research issues- demand characteristics

• refers to any cue from the researcher or research situation that may reveal the aim of the study and change participants behaviour

• if participants behave according to what they think the aim of the research is it means their performance on the tasks is likely to be artificial

• can be controlled by using a single blind procedure

research issues- investigator effects

• any effect of the investigators behaviour on the outcome of the research (the DV) and also on design decisions

• occur when the researchers presence or behaviour interferes with the research process and become a source of bias

• can be controlled by using a double blind procedure

ethical issues

who decides what is ethical and what isn’t?

ethical issues- deception

ethical issues- informed consent

ethical issues- privacy and confidentiality

ethical issues- protection from harm

peer review

aims of peer review- allocate research funding

aims of peer review- validation of quality and relevance of research

aims of peer review- suggest amendments or improvements

strength of peer review- protects quality of published research

limitation of peer review-

limitation of peer review-

limitation of peer review-

effects of psychological research on the economy

economical implications of attachment research into role of the father

economical implications of research into mental disorder treatment

reliability

ways of assessing reliability- test retest reliability

ways of assessing reliability- inter observer reliability

correlations and reliability

improving reliability- questionnaires

improving reliability- interviews

improving reliability- experiments

improving reliability- observations

validity

• the extent to which the research instrument measures what it sets out to measure, whether an observed effect is genuine and represents the real world (generalisable)

• data can be reliable but not valid eg an IQ test may produce the same result each time on the same people but not measure what its designed to

types of validity- internal validity

control within a study eg reduce demand characteristics, how much the findings of the DV have to do with manipulation of the IV