3.2.3.1 scientific processes

Hypothesis = a prediction about the outcome of results

Directional hypothesis = a one tailed hypothesis that states the direction of the difference or relationship

Non-directional = a two tailed hypothesis that does not predict the direction of the difference or relationship

Types of hypotheses:

Experimental hypothesis - there will be a difference

Null hypothesis - there will be no difference

Aim = the researcher needs to clearly state the intent of their research


Sampling:

Random sampling:

  • Choosing participants by a random method

  • e.g. names into a hat

  • Need a large pool of participants

  • Least biased

  • Equal chance of being chosen


Stratified sampling:

  • Subgroups of participants are created from the general population

  • P’s are chosen in numbers

  • More representative of the target population as all subgroup is represented

  • Each participant selected needs to agree

  • Time consuming

  • Can also be chosen at random


Systematic sampling:

  • A system to select participants

  • Numerical interval

  • Generally unbiased method as using an objective system

  • Not everyone can be picked


Opportunity sampling:

  • A group of P’s who have already been grouped

  • Therefore it is opportune to select them easily

  • Easy to obtain as it is ready for you

  • Biased

  • Quick


Volunteer sampling:

  • Where P’s self select themselves to take part in a piece of research i.e. newspaper

  • Wide access to a variety of P’s

  • Biased

  • Certain type of person just there to get paid

  • Volunteer bias = overly eager to participate


Pilot studies:

A small scale prototype of a study. Uses a smaller scale sample of the target population.

  • To check for any errors or problems i.e. instructions are clear, timings are correct, equipment needed is sufficient

  • Pilot studies are important to save time, money, resources


Experimental designs

Independent measures:

  • P’s take part in one level of the IV only. Using different groups of P’s in each condition

Strengths:

  • No order effects (practice/fatigue effects) because different P’s do each condition

  • Lower demand characteristics since each P does only one task

  • Same test can be used across all conditions as different P’s doing them

Weaknesses:

  • Participant variables (individual differences) may affect results as some may naturally perform better than others

—> to overcome, randomisation is used to balance out participants variables

  • Requires more P’s, making it more time consuming and harder to recruit


Matched pairs:

  • Pairs of P’s are matched in terms of key variables such as age and IQ. One member of each pair is then placed into the experimental group and the other member in the control group.

Strengths:

  • Participant variables are controlled because participants are matched on key characteristics

  • No order effects as one matched participant does one condition

  • Lower demand characteristics as one matched participant does one task —> less likely to work out aims

Weaknesses:

  • Matching process is time consuming, difficult, expensive —> affected by experimenter bias

  • Requires more P’s as each person is used once


Repeated measures:

  • The same P’s take part in each condition of the experiment. This means that each condition of the experiment uses the same group of P’s

Strengths:

  • No participant variables because the same people take part in each condition

  • Requires fewer P’s since each person is used multiple times

Weaknesses:

  • Order effect (practice/fatigue) can occur because the same P’s do all conditions - may get bored or tired —> perform better/worse

—> to overcome, counterbalancing occurs

  • Increased demand characteristics since P’s complete both conditions and may guess the aim


Observational designs

  • Structured and unstructured (behavioural categories)

  • Sampling procedures (time and event)


Behavioural categories = behaviour is operationalised into clear, objective behavioural categories

Structured:

  • The researcher has a list of specific behaviours they expect to see

  • Uses behavioural checklists

✅increases objectivity and reliability

❌some unexpected behaviour can be missed


Unstructured:

  • Researcher records all behaviour seen

✅uses pilot studies and when little known about behaviour

❌time consuming, difficult to analyse, some irrelevant behaviour can be recorded


Sampling procedures:

Event sampling: recording each time a specific behaviour occurs within a time period

Time sampling: recording behaviour at set intervals e.g. every 1 minute


Open and closed questions

Open questions

  • This is when the question is phrased in a way that allows the P to answer in any way they choose

  • e.g. “what do you think about X”

  • This produces qualitative data


Closed questions

  • This is when the question is phrased in a way that limits Ps responses to only a few fixed options

  • e.g. “do you like X, yes or no”

  • This produces quantitative data


Variables:

Extraneous variables = may have an effect on the DV and can affect the scores/behaviours in research. Usually environmental variables that cannot be controlled.

Situation variables = something with the situation/environment that affects the experiment i.e. a nuisance, an interruption

Participant variables = something to do with the P themselves meaning that they didn’t perform naturally in the research e.g. headache, stress


Overcome issues:

The single blind:

  • A control for demand characteristics

  • P’s do not know the aim of the experiment (or that they are in one) so that they don’t change their behaviour. This can cause deception/lack of informed consent so ethical issues are raised, and there are still opportunities for demand characteristics to take effect as the experimenter and the experimental situation can give cues

The double blind:

  • A control for demand characteristics and experimenter effects

  • P’s and experimenter don’t know the aim of the study. In this way it is thought that the experimenters will not have expectancies or transmit any cues to the P

Standardised procedures:

  • Helps to control experimenter effects

  • Experiments will often follow standardised procedures, such as set timings and materials to sure that all P’s have the same experience in the experiment. They will also give standardised instructions to each P

Counterbalancing:

  • Helps to control for order effects and fatigue effects

  • This is where the conditions/groups are randomised so that all P’s don’t do all the same conditions in the same order. e.g. AB, BA where one group does task A first and the other does task B first

Randomisation:

  • Helps control participant variables

  • Random allocation of P’s to groups to reduce the chance of individual differences (participant variables/bias)


Issues with experiments:

Demand characteristics:

  • P’s are unsure of how to behave and they try to work out what is require of them. This doesn’t show natural or everyday behaviours. Some P’s may actively try and mess up the data collection. This may become an extraneous variable and can affect the validity of the experiment - decrease the validity

Researcher bias:

  • Either consciously or not the researcher may impose a bias on the data collection. This will also be an extraneous variable and will affect the validity of the experiment = treat P’s in a different way

Order effects (practice/fatigue):

  • Doing the same task twice will produce a bias. Only occurs when the design is repeated measures. Also an extraneous variable and affects the validity of the data collection

—> all these issues need to be controlled if the validity of the experiment is going to be accurate



British psychological society (BPS)

  • Code of ethics and conduct - for professionals to protect P’s and integrity of research

  • Based on 4 principles: respect, competence, integrity, responsibility

Deception:

  • When information is deliberately withheld from the P’s or they are misled

  • This is unethical because it prevents P’s from giving fully informed consent

  • P’s must be debriefed and told the full reasons for deception and have the right to withdraw during the experiment

Right to withdraw:

  • Right of P’s to remove themselves from the study at any time - any data collected must be destroyed

  • It is unethical as it may cause the P’s stress if they feel forced to continue

  • P’s have the right to withdraw at the end of the study and must be fully debriefed

Privacy:

  • Right of P’s to decide how their information is used

  • It is unethical as researchers may obtain more information than necessary, this is an invasion of privacy, may lead to shame from the P

  • Keep information protected, don’t use P’s name

Confidentiality:

  • Personal information is protected by law

  • It is unethical because break the law if other parties access confidential information

  • Right to anonymity - P’s should never be identified, password protect information

Protection from harm:

  • P’s should be protected from physical, emotional or psychological harm (stress)

  • It is unethical because P’s should leave in the same state they entered

  • Offer right to withdraw, offer counselling, and fully debriefed

Informed consent:

  • When P’s consent to take part in a research - told the true aims

  • It is unethical because lack of informed consent means that P may not know what is fully required of them, it can cause stress and harm

  • Get ‘general’ consent, including consent to being deceived

  • Get ‘retrospective’ consent after the research is done


The scientific process of peer review

Peer review = process by which psychological research papers are subjected to independent scrutiny by other psychologists working in a similar field

Peer review is important as it provides a way of checking validity, making a judgement of the credibility assessing the quality and appropriateness of the design and methodology.

If a study was not peer reviewed it could potentially lead to the publication of invalid, or unscientific work.


Psychology and the economy

The implications of psychological research for the economy are concerned with how the knowledge and understanding gained from psychological research may contribute towards our economic prosperity.

If more effective treatments can be developed for psychological health problems, then this means that people will be a blessing to return to work and this reduces the burden on the employers, NHS, taxpayers.

Therefore, psychology can help shape and promote a better economy.


Test re-test reliability = doing the same test with different P’s and getting the same results

Inter-observer reliability = researchers independently give the same results to the same P’s at the same time - there is an 80% agreement

Validity = refers to wether something is true of legitimate (accuracy) - so the researchers knows exactly what is being measured

Internal validity = a measure of wether results obtained are solely affected by changes in the variable being manipulated in a cause and effect relationship

External validity = a measure of whether data can be generalised to other situations outside of the research environment

Face validity = a simple way of assessing wether or not something measures what it claims to measure

Concurrent validity = involves comparing a new test with an existing test to see if they produce similar results

Temporal validity = type of external validity that refers to the validity of the findings in relation to the progression of time

Ecological validity = a type of external validity that refers to the extent to which the findings can be generalised to a real life setting


Features of a science

Objectivity = data is free from bias or interpretation

Replication = confidence in research findings increase when investigations are replicated in the exact same way and the results are consistent

Empiricism = this evidence is gained through objective observations, experimentations, and the measurement of behaviour. Events are not distorted by emotions and are recorded as they happen

Falsification = part of the verification (validation process) is the idea of falsifiability, where a science theory or hypothesis must be empirically tested to see if it is false

Theory construction and hypothesis testing = a good theory is one that can be empirically tested. Unless you can test a theory, there is no means in knowing if it is a good or poor theory or if it needs modification. A good theory should therefore produce a number of testable hypotheses

Casual explanations = behaviour is caused by something other than free will. This assumption of determinism

Paradigm = a general theory or law that is accepted by a majority of scientists in a particular field of study. They are not fixed and are in changing

Paradigm shift = this occurs when enough evidence will accumulate so that the current paradigm is replaced


The scientific process of reporting experiments

Abstract = first section in a psychological report or journal, provides overview of full report, important to allow researchers to determine if the study is worth fully reading

Introduction = researcher reviews previous research to provide background info

Method = detailed description of methodology, needs to be detailed enough to allow someone else to replicate study

Results = quantitative, qualitative data, display of graphs/tables, inferential statistics

Discussion = where the researcher interprets the results of the study, makes criticism of the methodology, considers the implications of the results for future research

References = includes full details of any sources, important to correctly cite authors of studies to attribute their work to them