Scientific Processes

  • Created by: msahay
  • Created on: 15-08-19 12:53

Aims & Hypotheses

Aim - what is being studied. Hypotheses - precise, testable research predictions. 

  • Experimental/Alternative Hypothesis - predicts that differences in effects on the DV wil occur as a result of manipulation of the IV. e.g. 'alcohol consumption signficantly affects driving performance'. 

Experimental hypothesis is used when talking about experimental method, only. Other research methods use the alternative hypothesis - but the definition is the same.

  • Non-directional hypothesis (2-tailed) - predicts that there will be a difference in effects on the DV as a result of manipulation of the DV but does not predict the direction of the results. e.g. 'alcohol consumption signficantly affects driving performance'. 
  • Directional hypothesis (1-tailed) - predicts the direction of the effects on the DV as a result of manipulation of the IV. 'alchohol consumption signficantly worsens driving performance'. Used when there is availability of previous research or replicating a previous study. 
  • Null hypothesis - the manipulation of IV will not affect the DV. Any significant differences are due to chance factors. 
1 of 20

Sampling & Random Sampling

Sampling - the selection of participants to represent a wider population

Population - all people within a certain category e.g. schizophrenics, all 5-11 year olds etc.

From a population, we select a large, random, representative sample to be able to generalise our findings to the population as a whole. 

Random sampling - each member of a population has an equal chance of being selected. 

  • Unbiased selection - there is no bias in selection increasing the chances of getting a representative sample 
  • Generalisation - as the sample is fairly representative, findings will be generalisable to the whole population. 
  • Impractical - random sampling is difficult as sometimes not all members may be available or wish to take part. 
  • Not representative - unbiased selection does not guarantee a representative sample e.g. all females could be randomly selected 
2 of 20

Opportunity & Volunteer Sampling

Opportunity sampling - selecting participants that are willing to take part in the study

  • Ease of formation - opportunity samples are easy to create because ppts are readily available
  • Natural experiments - with natural experiments, opportunity sampling has to be used as the researcher has no control over who is being studied
  • Unrepresentative - sample is likely biased as only certain types of people are likely to take part e.g. those at work, if opportunity sampling was carried out at 12 noon on a weekday. Sample is not representative of the population as a whole, and findings cannot be generalised to the target population. 
  • Self-selection - participants have the option to decline to take part and the sampling technique thus turns into a self-selected sample. 

Volunteer sampling - people reply to an advert posted by researchers for volunteer ppts.

  • Ease of formation - volunteer sample is easy to create as participants volunteer themselves.
  • Less chance of screw you phenomenon - participants are eager to take part and there will be less chance of them trying to sabotage the study deliberately. 
  • Unrepresentative - volunteers are a certain type of people/demand characteristics
3 of 20

Systematic & Stratified Sampling

Systematic sampling - selecting every nth person from a list of population

  • Unbiased selection - there is no bias in selection, increasing chances of getting a representative sample. + Generalisation - the results are representative of the population which means that findings can be generalised
  • Periodic traits - the process of selection can interact with a hidden periodic trait within the population e.g. if every 5th property in a street is owned by a young person, selecting participants who live at every 5th property is not representative + Not representative - unbiased selection does not guarantee an unbiased sample making the sample somewhat unrepresentative and findings difficult to generalisable. 

Stratified sampling - dividing a population into stratums relevant to research, then randomly sampling within each stratum. 

  • Representative - selection occurs from representative sub-groups within a population so sample should be fairly representative + Unbiased - random sampling is performed upon the subgroups meaning selection is unbiased sample 
  • Knowledge of population characteristics required - detailed population characteristics needed for stratified sampling may not be available + Time-consuming - dividing and then randomly selecting can be time-consuming 
4 of 20

Pilot Studies

small scale version of the main study that is conducted to identify any problems in the method, design or measurements. 

They ensure that:

  • the procedures of the study run smoothly 
  • equipment/tests are functioning properly
  • participants understand instructions
  • extraneous variables are controlled
5 of 20

Experimental Design: Independent Groups

participants are split into 2 groups, each group performing 1 condition only (either the experimental or control condition)


  • No order effects - the order in which conditions are performed does not have an effect on the outcome as each participant performs only one condition of the experiment. 
  • Time saved - both sets of participants can be tested simultaneously. 


  • Participant variables - individual differences between participants in the two conditions may affect the DV, rather than the IV manipulation, reducing validity of results.e.g. participants in one condition may be more intelligent than those in another condition.
  • Can be addressed by random allocation of participants to the conditions of the IV and a large sample 
  • More participants needed than a repeated measures design - as each participant only performs one condition 2x the number of participants are needed. 
6 of 20

Experimental Design: Repeated Measures

all participants perform the 1st condition, followed by the 2nd condition. This allows us to directly compare participants' performance across the 2 conditions. 


  • No possibility of participant variables - threatening the validity of the study as the same people are measured in all conditions. 
  • Fewer participants are needed - as each participant performs both conditions and gives two sets of data. 


  • Order effects - as participants perform all conditions, the order in which they perform these may impact the results and their validity e.g. participants may improve due to practise or get worse due to boredom. 
  • Counterbalancing can control the above - 1/2 of participants perform condition 1, then 2 whilst other 1/2 perform condition 2, then 1. 
  • Demand characteristics - as participants perform all conditions, they may be more likely to guess the aim of the study and exhibit demand characteristics by acting accordingly. 
7 of 20

Experimental Design: Matched Pairs

participants are in similar pairs (e.g. similar age), with 1 of each pair performing 1 one condition only. 


  • No order effects - each participant performs one condition only, there are no order effects. 
  • Participant variables - as participants are matched, the effect of individual differences between participants on the DV should be reduced. 


  • More participants needed - because each participant only performs 1 condition. 
  • Matching difficulties - matching participants on all variables is impossible and an unmatched variable may be vitally important, making it a time consuming process. 
8 of 20

Ethical Issues

the rules governing the conduct of researchers in investigations

Informed consent - researchers must give participants sufficient details about a study so that they can make an informed choice as to whether they want to participate or not. 

Avoidance of deception - withholding information or misleading participants is unacceptable if participants are likely to object or show unease once debriefed. If deception is used, informed consent cannot be gained from participants. Often, however, a study uses an element of deception so that participants do not guess purpose of study.

  • Presumptive consent - gained from people of a similar background to the actual participants in a study. If they state that they would have been willing to participate, then it is deemed the actual participant would have as well. 
  • Prior general consent - participants agree to be deceived without knowing how they will be deceived, but this may affect their behaviour. 
  • Retrospective consent - asking participants for consent after they have participated in a study. However, they may not consent and have already takn part. 
9 of 20

Ethical Issues Continued

Adequate briefing and debriefing - all details of a study should be explained to participants before and after. Debriefing must be done if deception is used. 

Protection of participants - researchers have a responsibility to protect participants from physical or psychological harm. 

Right to withdraw - participants should be regularly reminded that they can leave the study whenever they wish and can withdraw their data after the study is completed. 

Confidentiality/anonymity - participants' data should not be disclosed unless agreed in advance and their identities e.g. names should be replaced. 

Observational research - observations are only made in public places where people might be expected to be observed by strangers

Incentives to take part - participants must not be offered money as this puts pressure on them to take part. 

10 of 20

Psychology Research Effects on Economy

Research contributes to the economy by creating effective therapies for mental disorders.These will save money as many people will be able to return to work and contribute to the economy by doing their job. 

Psychology assists in finding solutions to wider social problems relating to crime, aggression, child abuse etc..

This could contribute to the economy by reducing levels of crime, reducing prison population and increased taxation (people working rather than being in prison) 

11 of 20


the extent to which a test or measurement produces consistent results

If a study is repeated using the same method, design and measurements and the same results are obtained, the study is said to be reliable as it indicates a cause-effect relationship of the manipulation of the IV on the DV. 

  • Internal reliability - the extent to which something is consistent within itself. 
  • External reliability - the extent to which a test measures consistently over time. 

Split-half method - splitting a test into 2 and having same participant complete both halves. If the two halves produce similar results, the test has good internal reliability. 

Test-retest method - giving the same test to the same participant on 2 separate occasions. If the same result is obtained, the test has good external reliability. 

Inter-observer reliability - means of assessing whether independent observers are observing and recording behaviours in the same way with behavioural categories being clearly defined before the study. Analysing the correlation between different observers' scores on different behaviours can be used to assess inter-observer reliability - a high score indicator indicates good IRR. 

12 of 20


how accurately a study investigates what it claims to (internal) and the extent to which findings can be generalised beyond the research setting (external). 

Internal validity - whether results are due to the manipulation of the IV and not affected by extraneous cofounding variables. 


  • reducing investigator effects 
  • reducing demand characteristics
  • standardised instructions
  • random sample 

External validity - the extent to which findings of the study can be generalised to other settings (ecological validity), other people (population validity) and over time (temporal validity). 


  • naturalistic settings
13 of 20

Ways of Assessing Validity

Face validity - assesses extent to which items look like what a test claims to measure

Concurrent validity - assesses validity by correlating scores on a test with a known valid test

Predictive validity - assesses validity by predicting how well a test predicts future behaviour, e.g. school tests predicting later exam results. 

Temporal validity - assesses extent to which research findings remain true over time. 

14 of 20

Features of Science

The scientific process - a means of acquiring knowledge based on observable, measurable evidence. 

1. Observation of a phenomenon

2. Formulation of a hypothesis to explain the phenomenon

3. Performance of proper experimental tests of the predictions by researchers

  • Empiricism - information is gained through direct observation or by experiment on physically observable, measurable phenomena rather than by thoughts or beliefs. 
  • Objectivitiy - observations are unbiased and non-interpretative.To lessen unconcious bias researchers use standardised instructions, operational definitions of variables + peer review. 
  • Replicability - being able to repeat a study to check its validity. The design + methodology of the study must be written up carefully so other researchers can check its validity. 
  • Falsification - the idea that scientific statements are capable of being proved wrong - replication is a way of determining this. Statements must be empirically testable to be considered scientific. 
  • Control - researchers aim to demonstrate causal relationships between variables. 
15 of 20

Theory Construction & Hypothesis Testing

Inductive phase - observations give information used to formulate theories about the phenomena

Deductive phase - predictions from theories known as testable hypotheses -->experimentally tested --> their results analysed --> adjustments are made to the theory

Laboratory experiments are the most scientifically empirical as they allow us to establish causality.

Other methods like field and natural experiments reduce the capability to determine causality. But using objectivity improving methods such as inter-observer reliability that attempts to ensure an unbiased observationfindings can be claimed to be valid. 

Paradigma shared set of assumptions about the subject matter. 

Paradigm shiftsrevolutionary changes in scientific assumptions 

Kuhn believed that scientists create a bias where they try to find confirmatory results and not non-confirmatory examples. 

16 of 20

Reporting Psychological Investigations

  • Title 
  • Table of contents
  • Abstract
  • Introduction - why the study was conducted (previous research investigations in chosen topic)
  • Aims - what the researchers aim to investigate 
  • Hypotheses - the experimental/alternative hypotheses and the null hypotheses are stated and level of significance. 
  • Method - outline of what was done in the study so it can be properly replicated e.g. standardised instructions

Design, participants, apparatus, standardised procedure/instructions and brief/debrief instructions

  • Findings - what was found in terms of data collected

Statistical data including descriptive statistics (tables, averges and graphs) and inferential statistics (the use of statistical tests to determine how significant the results are). 

17 of 20

Reporting Psychological Investigations 2

If you are asked to outline and discuss the results of a study: 

  • Write the results clearly in words e.g. the mean number of items remembered for participants listening to music was 7, but for those not listening to music, it was 9. 
  • Refer to the standard deviation/range and explain what they mean e.g. those listening to music had a higher standard deviation than those not listening to music, meaning that their scores varied more around the mean. So there were more individual differences in participants' memories when listening to music 
  • Say whether the results were significant and how you know this (refer to the OV, CV and level of significance) and what it means
  • Discuss issues of validity + reliability 
  • Discussion - explanation of results, relationship to background research, limitations/modifications, implications and suggestions for future research 
  • Conclusion + references
18 of 20

Peer Review

scrutiny by experts of research papers to determine scientific validity before research is published. 

Peer review provides a way of:

  • checking the validity of the research
  • making a judgement about the credibility of the research
  • assessing the quality and appropriateness of the design and methodology.

which helps to prevent incorrect and invalid data entering the pulic domain. 

Peers can suggest whether the paper should be published, revised or rejected in some way which helps to ensure that any research paper published in a well-respected journal can be taken seriously by other researchers and the public. 

  • Single blind review - names of reviewers not revealed to researcher as reviewer anonymity allows for unbiased review. 
  • Double blind review - both reviewers and the researcher are anonymous. Researcher based bias is not present by the reviwers which allows for unbiased review 
  • Open review - reviewers/researchers know eachother. Reduces plagiarism/personal risk
19 of 20

Peer Review Criticisms

Single blind review - anonymous reviewers may delay the reviewing process so that they can publish similar research first. 

Double-blind review - the researcher can probably be identified by their writing style 

Open review - deserved criticism is watered down by reviewers in fear of being polite

  • Peer review isn't actually unbiased as research occurs in a narrow environment - relationships between researchers working in similar areas affects their impartiality and objectivity.
  • Obscure research areas may not have reviewers with specific knowledge to carry out a proper peer review. 
  • Researchers are funded by organisations that only want certain research to be seen as scientifically acceptable, compromising their ability to be unbiased
  • Plagiarism/publishing first problems
  • Peer review is slow 
  • Peer review is controlled by elites that may be resistant to revolutionary ideas in research
  • False/unscientific research being accepted as true is a real problem because many researchers can base their own findings on this. 
20 of 20


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »