Headstart | Research Methods

aim

a statement outlining the purpose of the investigation

hypothesis

a testable prediction of the relationship between two or more variables

theory

a body of interrelated concepts (‘ideas’) that attempt to explain interrelated observations and make predictions about future events

model

focuses more on representing how some behaviour and/or a mental process(es) could, should or does occur

variable

something in which individuals, animals or objects differ among themselves, can change (vary) in amount or kind, and is measurable

independent variable

manipulated by the experimenter to measure the effects of the DV

dependent variable

the participant’s responses which are measured to assess the effects of the IV

operationalising the IV & DV

involves defining how they will be manipulated or measured in the experiment

benefits of operationalising the IV & DV

ensures the IV and DV are testable and therefore that the research hypothesis is testable. Helps avoid experimenter biases and differences that can affect the results in an unwanted way.

controlled variable

a variable that is considered to have an effect on the dependent variable so it is held constant to remove its potential effect

extraneous variable

any variable other than the IV that may influence the DV and affect the results in an unwanted way

confounding variable

an uncontrolled and unwanted variable that changes with the IV

population

the larger group of research interest from which the sample is drawn

sample

smaller representative subset of participants drawn from the population of research interest

representative sample

a sample that closely resembles the population from which it is drawn in key characteristics

biased sample

a research sample that does not adequately represent the key characteristics of its population

law of large numbers

that as sample size increases, the attributes of the sample more closely reflect the attributes of the population from which the sample was drawn

stratified sampling

dividing the population of interest into subgroups based on characteristics (strata) such as age, sex, religion, wealth, ethnicity, etc.

stratum

a separate sample from each subgroup (‘stratum’) is then selected so that they are represented in the sample in the same proportion that they were in the population

steps for stratified sampling

STEP 1: identify the population of research interest

STEP 2: select the strata (eg. gender)

STEP 3: work out the proportion (percentage) of the population that fits each strata (eg. males and females) STEP 4: work out how many of each strata you need in the sample that is in the same proportion as the population STEP 5: randomly select the number of participants from each strata required for the sample

convenience sampling

participants selected on the basis of being readily available – may produce a biased sample / not able to generalise.

between subjects design / independent groups

each participant is assigned to only one group or condition and provides only one score for data analysis

within subjects design / repeated measures

each participant is in both the experimental and control groups or all the treatment conditions

mixed design

combines features of both a between subjects design and a within subjects design. This means that the researcher can assess the potential differences between two or more separate groups of participants

advantages of between subject design / independent groups

* usually do not have to spread out the time period between experimental conditions; can be done in one go, preventing participant dropout
* no order effects

disadvantages of between subject design / independent groups

* a larger number of participants are required to help ensure a larger variety of participant variables
* less control over participant variables

advantages of within subjects design / repeated measures

* can control the unwanted influence of variables
* requires a smaller number of participants

disadvantages of within subjects design / repeated measures

* does not necessarily control all participant variables that could influence the results
* fatigue is more likely to occur

advantages of mixed design

* the researcher can capitalise on the strengths of the between subjects and within subjects designs
* more precise and detailed responses as fewer participants but greater sensitivity in results

disadvantages of mixed design

* more complex to complete as requires more expertise to collect and analyse the data than other methods
* extra resources as combining two methods:
* time
* money

random sampling

representative sample as every member of the target population has an equal chance of being selected for the study

advantages of random sampling

* useful for comparing specific subgroups within populations
* increases sample representativeness

disadvantages of random sampling

* requires access to a relatively complete and up-to-date list of target population and contact details to ensure representativeness

advantages of stratified sampling

* ensures selection of a representative sample
* minimises sample bias in relation to participant variables

disadvantages of stratified sampling

* time consuming
* resource intensive

advantages of convenience sampling

* time and cost effective
* easy participant access

disadvantages of convenience sampling

* tends to produce sample bias
* not very representative of the population
* limits generalisations

research method

a particular way of conducting a research study

controlled experiment

an investigation of the relationship between an iv and a dv, whilst controlling all other variables

* this can establish a cause-effect relationship if well designed

correlational studies

investigates the relationship that exists between variables without any control over the setting in which the relationship occurs or any manipulation by the researcher

* correlation ≠ causation

positive correlation

the relationship between two variables that change in the same direction

negative correlation

the relationship between two variables that change in different directions

zero correlation

there is no relationship between the variables

quantitative correlation

by a correlation coefficient that ranges from -1 (strong negative) to +1 (strong positive)

correlation advantages

* used to test hypotheses where it is not desirable/possible to manipulate the IV of interest
* can conduct outside an artificial laboratory situation where results may also be more realistic
* useful for discovering relationships between variables
* correlations can indicate patters or trends; contributing to the development/testing of theories/models

correlation limitations

* does not permit the researcher to draw firm conclusions about cause-and-effect relationships
* not a legitimate research method for studying causality

self reports

a participants answers to questions presented by the researcher

* e.g questionnaires, interviews
* can be free response or fixed response questions

structured interview

the participant is asked specific, pre-determined questions in a controlled manner

* e.g. interviewer uses a script, neutral manner, no facial expressions or cues

semi-structured interview

interviewer may use free response questions, but still using a script to ensure participants are all treated equally

unstructured interview

researcher has an overall aim of what data should be collected but spontaneously generates questions for each participant

questionnaire

a written set of questions or other prompts designed to draw out self-report information from participants on a topic of research interest

* e.g. surveys to large populations

questionnaire advantages

* time efficient
* anonymity of participants
* can survey a large number of participants

focus groups

a small set of people, typically 8-12, who share characteristics and are selected to discuss a topic of which they have personal experience.

leader is present to conduct discussion, keeps on target and encouraging free-flowing, open-ended debating

advantages of self reports

* able to collect data from a large number of people in a short period of time
* cost effective
* easy to administer
* easy to compare to create a study
* anonymity

limitations of self reports

* assumption that people are self aware of their behaviours
* honesty
* bias
* social desirability

social desirability

giving false or misleading answers to create a favourable impression of themselves

observational studies

involves a collection of data by carefully watching and recording behaviour as it occurs without any intervention or manipulation

naturalistic observation

the researcher watches and records behaviour in the natural, ‘real life’ environment where it would ordinarily occur without manipulation of variables or other controls that would occur in a laboratory setting

contrived environment

one that the researcher creates or sets up for the specific purpose of conduction an observational study

* artificial, ‘non-naturalistic’, controlled

participant observation

a trained investigator studies a pre-existing group by joining it as a member, while avoiding a conspicuous role that would change what occurs in the group and bias the data.

* researcher role may be known or unknown to other members

non-participant observation

the researcher tries to conceal their presence so that their observations are made in an entirely inconspicuous manner

advantages of observational studies

* naturalistic observation - can watch/record everyday behaviour without the need for manipulation/intervention
* can use checklists/specific criteria to enhance accuracy
* some behaviours can only be studied as they occur naturally
* could be unethical/impractical to study in a laboratory setting
* does not required the cooperation of the participants being observed

limitations of observational studies

* cannot be used to determine the cause of behavioural interest that is observed, because many factors may influence that behaviour
* lack of control of such variables
* naturalistic observation studies often lack a representative sample
* observer bias

case studies

an intensive, in-depth investigation of some behaviour, activity, event or problem of interest in a single individual, group, organisation or situation

advantages of a case study

* depth of analysis
* richness of data
* can avoid artificiality and provide a ‘snapshot’ of the actual or real life experience of one or more individuals conducted over a long time period
* provides insights into how others may think, feel or behave under similar circumstances
* a valuable source of hypotheses for further research or for data to support theory building or challenge a theory’s assumptions

limitations of a case study

* cannot test or establish a cause-effect relationship as does a controlled experiment
* small sample size
* lack of generalisability
* time consuming
* often susceptible to biased information from the participants or researcher

simulation studies

reproducing situations of research interest in a realistic way to investigate the behaviour and/or mental processes of individuals in that environment

* e.g. stanford prison experiment

advantages of simulation studies

* can be suitable alternative when the real environment is not available or possible
* diversity of research questions that can be investigated means that they can be a valuable source of hypotheses for further research
* when a simulator and/or computer assisted technology is used, this may also be time and cost effective
* greater experimenter control
* the capability of reproducing an identical environment whenever needed also supports replication of experiments to test the results

limitations of simulation studies

* artificial and therefore studies may lack realism
* makes difficult to generalise the results

random errors

errors due to some chance factor or variation in a measurement

* e.g. a participant feels sick during an experiment and this impacts results

systematic errors

produced by some factor that consistently favours one condition rather than the other

* e.g. a faulty measuring instrument that repeatedly gives out a false reading each time

the difference between random and systematic errors

random errors affect the precision of a measurement, whereas systematic errors affect the accuracy of a measurement

accuracy

relates to how close it is to the true value of the quantity being measured

precision

refers to how closely a set of measurement values agree with each other

personal errors

a fault or mistake by the researcher

types of extraneous variables

participant variables, situational variables, demand characteristics, experimenter effects and the placebo effect

placebo

inactive or fake substance or treatment

placebo effect

a change in the participants response or behaviour due to expectation

* **solution**, double blind procedure | participants are unaware to which group they have been allocated

experimenter effect

unintentional personal characteristics of the experimenter can affect participant responses

* **solution**, double blind and triple blind procedure

double blind procedure

participants and experimenter are unaware of which group they have been allocated, however a researcher would use an assistant to give out the medicine who is aware

triple blind procedure

experimenter, research assistants and participants are all unaware of who has been allocated each group

participant variables

the personal characteristics that individuals bring to an experiment that could influence their responses

* could be biological, psychological and/or social in nature

situational variables

external factors (other than the IV) associated with the experimental setting that may influence participant responses

* e.g. background noise, lighting, experimenter and research assistants, research activities and their sequence

order effect

occurs when performance on the DV is influence by the specific order in which the experimental tasks are presented rather than the IV

* may change the results so the impact of the IV may seem to be greater or less than it really is

practice effects

influence on performance (the DV) that arises from repeating and/or prior experience with a task, including the test materials, procedures and settings

carryover effects

influences that a particular task has on performance in a task that follows it

* arises simply from experiencing a task

controlling situation variables

* hold them constant throughout the experiment
* balancing or equalising their effects for all groups of participants
* randomisation
* testing participants in one condition first, then all participants in the other condition
* standardised instructions and procedures
* counterbalancing order effects

demand characteristics

cues in an experiment that may influence or bias a participant’s response, thereby distorting the results

* mostly intentional or unconsciously done

beneficence

the commitment to maximising benefits and minimising the risks and harms involved in taking a particular position or course of action

integrity

the commitment to searching for knowledge and understanding, and the honest reporting of all sources of information and results, whether favourable or unfavourable, in ways that permit scrutiny and contribute to public knowledge and understanding

justice

moral obligation to ensure that there is fair consideration of competing claims; that there is no unfair burden on a particular group from an action

non-maleficence

avoiding the causations of harm; however, as a position or course of action may involve some degree of harm, the harm resulting from any position or course of action should = or > than the benefits

respect

involves consideration of the extent to which living things have an intrinsic value and/or instrumental value; giving due regard to the welfare, liberty and autonomy, beliefs, perceptions, customs and cultural heritage of both the individual and the collective

ethical concepts

supplemented by the following ethical guidelines, which should be considered when conducting and evaluation psychological investigations

confidentiality

the privacy, protection and security of a participant’s person information

debriefing

at the end of the experiment the participant leaves understanding the experimental aim, results and conclusions

informed consent procedures

ensuring participants understand the nature and purpose of the experiment; including potential risks, before agreeing to participate in the study.

voluntary written consent should be obtained by the experimenter by either the participant or a parent/legal guardian if the participant cannot give consent

use of deception in research

only permissible when participants knowing the true purpose of the experiment may affect their behaviour while participating and could in turn effect the validity of the experiment

voluntary participation

ensures that there is no coercion or pressure put on the participant to partake in the experiment

withdrawal rights

a participant is able to discontinue their involvement in the experiment at any time during or after the conclusion, without penalty.

* e.g. removing participant results