Scientific Inquiry

Aim

a general statement explaining the purpose of research

Independent variable

Variable being manipulated by the researcher

Dependent variable

Varibale being measured by the researcher

To develop research, what do you need?

A testable hypothesis

Operational definitions of both independent and dependent variables

Controlled Variables

Variables that stay consistent throughout an experiement

Extraneous variables

Unwanted factors that may impact the dependent varibale, and researchers may not be aware of them until after the study is completed

• when it is identified and managed, it becomes a controlled varibale

Participant variables

what are they and how to manage them

examples

a type of extraneous variable related to the individual characteristics of each participants

can manage by

• selecting participants with similar characteristics suitable to the study

• use random allocation to ensure equivalent groups are created

examples

• motivation, educational background, age, gender, self-esteem, memory, prior experience, health, mood

Environmental variables

extraeous variables related to the enviroment the study takes place in and hoow it may influence participant responses

• testing venue

• background noise

• room temp

• time of day

researcher variable

Extraneous variables that are associated with the personality characteristics, appearance, conduct of the researcher that can untentionally impact participant responses

• accent

• gender

• attractiveness

• health

• age

• interactions with participants

Confounding variables

that impact the dependent variable and also have a causal/correlational relationship with the independent varibale

• alter relationship between IV AND DV, so can complicate results + make them difficult to interpret

• uncontrolled extraneous variables can becoming confounding

• e.g stress has correlational releationship with IV. stress affects DV as higher stress can increase time taken to fall alseep

directional hypothesis

statement that compares the predicted outcome of each condition

it is hypothesised that students will take less time to do blah blah blah

non-directional hypothesis

states that a diffrence exists but does not specify the nature of the conditions

• it is hypothesised that students who take hemp seed oil before bed will differ in time taken to fall asleep than…

inquiry question

does not predict outcomes but prompts broad exploration of research topic

• mostly used in qualitative data formation

• shapes overall methodology

• e.g will hemp seed oil decrease the time it will take to fall alsleep

Experimental research

manipulating the independent variable to determine its effect on the depedent variable

allows researchers to establish cause and effect

participants randomly allocated to groups

• experimental group: exposed to IV

• control group: basis for comparison with experimental group, enabling researchers to determine whether IV has affected DV

e.g harlow’s use of infant rhesus monkeys to test comfort>food

Strengths of experimental

• allows researchers to have control over variables, minimising influence of extraneous varibales

• enables identification of cause and effect relationships between IV and DV

Limitations

Having a controlled environment such as lab environment reduces realism (lowering external validity), may impact participant variables

in trying to control variables, there’s risk of human

Non-experimental research

studies where the IV cannot be manipulated

participants cannot be randomly allocated

cause and effect relationships cant be established

• case studies, observational research, correlation studies

Strengths of non-experimental

allows for observation of naturally occuring behaviour without need for controlled setting

useful in studying situations where manipulating varibales would be unethical

Limitations

cannot provide reliable causal conclusions as it doesn’t establish cause and effect

no variable manipulation, so larger sample sizes required so more participants are able to be observed

Types of non-experimental research

Observational

Case study

Correlational study

Observational

type

application

method

strengths

weaknesses

non-experimental

• type of technique used to study behaviour

• method: researchers monitor participants and record notes

• strengths:

  • controlled observations can be replicated,

  • more likely to behave naturally rather than consciously/unconsciously acting in a socially appealing way

• limitations:

  • researcher sees what they expect to see or recoreds selected details,

  • observer bias can occur, participants may change behaviour if aware of being observed (observer effect),

  • voluntary participation and infomed consent may be breached

Case study

type

application

method

strengths

weaknesses

non-experimental

application

• in depth investigation of individual, group, single event that are useful for examining unusual effects that cannot be replicated

method

• large amount of data (mostly qualitative) is collected,

• providing info on one person, group of people, or event

strengths

• detailed information

• range of perspectives

weaknesses

• results can’t be generalised to population the sample was taken from

• conclusions drawn from case studies are limited due to lack of formal control groups

Correlatonal

type

application

method

strengths

weaknesses

non-experimental

• measures the linear relationship between two variables (called co-variables)

  • e.g positive correlation was found between birth weight and intelligence in a 2020 study of 1719 children from danish national british cohort

method

• relationship between two co-variables is measured

strengths

• potential hypothesis based on correlation can be tested using experimental design

• can be used when manipulating variables in experimental research is unethical

limitations

• correlations do not show how variables are related because there is no cause and effect between two variabels (correlation does not infer causation)

• extraneaous varibales are not controlled and could invervene with the relationship between variables, making it hard to know if the relationship would’ve existed otherwise

naturalistic observation

researchers observe participants in their natural setting in an unobtrusive manner

advantages

• high external validity

• suitable for studying concepts not able to study in a laboratory

limitations

• observer effect (awareness of being watched causes participants to alter their behaviour and observer bias where the observer’s expectations or beliefs influence what they record during observational research)

controlled

researchers observe participants in an environment that is structured, such as lab

strengths: increased accuracy in observations due to greater environment control

limitations: participants alter behaviour due to being watched

research study designs

longitudinal

cross sectional

longitudinal

application

method

strengths

limitations

application

• used to study developmental trends across the lifespan

method

• data collected more than once, using same participants (weeks, several days, years or decades)

strengths

• developmental trends can be studied over lifetime

• frequency/timing or duration of events can be assed (depressive symptoms)

limitations

• it takes a longer to get results than cross-sectional studies

• participants may drop out of the study along the way

correlation

application

method

strengths

limitations

application:

• often used to determine prevalence of diseases or health conditions in a community

• useful for population based surveys

method

• data from participants is collected at one point in time

• may be from one sample or several

strengths

• quicker than longitudinal, no follow ups needed

• costs less

limitations

• results may differ if another time to collect data was chosen (snapshot in time)

• sample size may not be large enough to generalise results to population

population

the entire group of people that is of interest to the researcher

sample

subsection of the population

important that sample is representative of the population it was taken from, allowing generalisability

sampling

process of selected participants from population of research interest to participate in a study

sampling methods

convenience

snowball

random

stratified

convenience

researchers choose participants who are convenient for them to reach

• easily accessible participants

• employed by teachers or university professors who study their own students

strengths

• less time + effort to collect sample than random or stratified

• costs are lower

limitations

• researcher bias

• unlikely to be representative of population, lowering generalisability

snowball sampling

application

method

strengths

limitations

initial participants are chosen and then each participant encourages other ppl to contact the researcher and join sample

• allows access to groups of participants who are otherwiae not easily identifiable such as homless and minority

strengths

• allows researchers to find a sample that may otherwise be difficult to recruit due to nature of the study (e.g sex workers)

• time needed to gather sample is reduced because initial participants recruit more

limitations

• unlikely to be representative of population since researchers are minimally involved in participant recruitment

• sample may be biased as researchers can only recruit those who are in direct contact with OG

random sampling

application

method

strengths

limitations

allows every person in a sample to have an equal chance of being randommly selected to be a member of the sample

method: compiling the names of all members and randomly selecting them

applications: educational studies (e.g student performance and healthcare research)

benefits:

• researchers do not personally select participants in their sample

• equal chance of being selected

limitations:

• significant amount of time and effort to conduct effectively

• sample size not large enough = may not be representative

stratified sampling

application

method

strengths

limitations

application: educational studies

requires population of interest first be broken down into sub groups based on characteristics relevant to the study then participants from each subgroup are randomly selected in the same proportions they appear in the population

strengths:

• sample is more likely to be representative of the population

• researcher bias is minimised since participants not handpicked by researcher

limitations:

• significant time and effort required to conduct sampling process effectively

• researchers may not always be able to classify each participant of population into subgroups

Random Allocation

involves the random distribution of participants into experimental and control groups to reduce selection bias and increase generalisability

1) names of participants in sample collated

2) names selected randomly

strengths

• good for generalisability bc equivalent groups or participants are created

• prevents selection bias because each participant has equal chance of being placed in different conditions

limitations

• inability to use the method for cases where IV cannot be manipulated

• does not guarantee groups are equivalent in terms of characteristics

difference between random sampling and random allocation

random sampling: method used to select participants from the population to join the sample

random allocation: separation of participants from the sample into groups

sources of extraneous variables

experimenter effect

demand characteristics

experimenter effects

how does it occur

what does it do

occur when researcher’s expectations or behaviours bias results

give away desired outcome or unintentionally influence participants

behave in a certain way that participants interpret as clues for how to behave

presenting different instructions to different groups + innacuraly record/interpret data basde on expectations

how to reduce experimenter effects

double blind procedure

• where researcher + particpants are unware of conditions (do not know which condition of the IV participants are allocated to)

• reduces experimenter effects bc researcher does not know which are exposed to IV, so less likely to unintentionally influence participants by treating them differently based on their group

demand characteristics

cues participants perceive during a study that lead them to believe they have discovered the aim of the study or expectations of the researcher

• causes them to behave (often unconsciously) in ways that support the hypothesis or help achieve what they believe to be the desired results

• may not come from methodology/researcher behaviour but from rumour/location of lab

• can occur to please the researcher or be viewed positively by them

• can cause participants to purposefully want to disprove hypothesis + ruin credibility

• even if researchers do everything, participants can still believe they’re discovered aim and expectations

can the experimenter effect allow for demand characterstics

researcher unintentionally sharing expectations of study may lead some participants to believe they now know the researchers desired outcomes and change behaviour to help create them

how can the chance of experimentor effects occuring be reduced?

single or double blind procedure

single blind

researcher is aware of the aim and experimental conditions (which are in CG and EG) while participants are unaware

• participants often not told aim

• knowing aim can affect behviour and result in extraneous variables such as experimenter effects and demand characteristcs

• deception is used in such cases + participants would have the true purpose of study and reason for deception explained in debriefing

effects of extraneous and confounding variables

placebo effect

how to prevent demand characteristics?

• single blind/double blind can help

• as well as placebo

placebo

neutral treatment that looks the same as the real treatment being evaluated and is delivered in the same way

• can minimise demand characteristics bc it limits participants from knowing true nature of experimental condition, thus preventing them from discovering aim of research

• less likely to alter behaviour + support researcher expectations if they do not know if they are receiving placebo or actual treatment

placebo effect

positive result that occurs due to participants belief that a treatmment will be effective

consequence of experimenter effect and demand characteristics

this subconcious alteration of behaviour to align with their bleif can arise as a reesult of extraneous variables like demand characteristics and psych facotrs

what can reduce effects of extranous and confounding variables?

random allocation

placebo

single + double blind procedures

standardisation of procedures and instructions

random allocation purpose

to ensure each participant in the sample has an equal chance of being chosen for the control group as for the experimental group

how can random allocation reduce participant variables?

ensures participants with various personal characteristics are spread between experiemtnal and control groups,

researcher wants to find out if changes to DV are due to IV, and not due to personal charactestics

how can random allocation reduce researcher effects

preventing the researcher from being able to distribute participants into groups and personally decide who is going to be in the experimental group

standardisation of procedures and instructions in minimising environmental variables

by providing the same location and conditions for all participants, for example, conducting an experiment in a lab setting, providing same instructions to each group of participant can minimise researcher variables and experimenter effects

how can confounding variables be avoided

controlled variables

qualitative data

descriptive information in the form of words

quantitative data

information in the form of numbers that can be counted

qualitative data collection techniques

interviews

open ended surveys

• both subjective measures that require participants personal perceptions and interpretations of their experiences

interviews

self report, where researcher asks participants questions in real time (face to face, over the phone)

structured interviews: set of pre-est questions asked individuals or focus groups in real time

strengths of interviews

many indv+ groups can be asked same set of standardised questions, reducuing differences between intervieweers

participants do not need to rely on reading ability

limitations of interviews

unable to ask participants to further explainr responses, limiting richness of collected info

analysing data collected from iinterviews is complicated so drawing general conclusions is hard

semistructured interviews

set of pre-est Qs asked in real time but participants can additionally be asked follow ups based on early response (e.g j*b interview)

strengths of semistructured interviews

extensive data can be collected

option to ask further questions leads to deeper understanding

participants do not need to rely on reading ability to participate

limitations

• face to face, so may feel uncomfy with revealing sensitive info to interviewer, limiting data collection

• analysing data can be complicated making it difficult to draw general conclusions

open ended survey

participants are provided with question son paper or online space to respond in open text format with as much detail they want

used in exploratory studies of issues requiring deep insight

strengths

detailed info such as attitudes, emotions can be collected on complicated topics

• not restriced by limiting options such as rating scales

• completed online or posted anywhere allow geographical accessibility

• prevents geographical barriers and allows broader rep in research

limitations

must rely on reading capabilities and writing ability

differences in detail provided by them makes data analysis difficult

quantitative objective

objective physiological measureds that do not rely on personal interpretations or perceptions.

information based on facts that can be supported through observation

examples

heart rate, galvanic skin response

heart rate

changes in emotional stress, physical effort and conciousness can be recorded by measuring heart rate or breathing rate of participants

• recorded via bpm using an electrocardiogram that records electric signals in the heart

breathing rate

measured by number of breaths per minute and can be measured with or without equiptment

can detect changes in emotional stress and physical effort and concioussness withheart rate too

galvanic skin response

determines changes in electrical conductivity of the skin

• can detect anxiety, guilt, fear, excitement

• can be used to determine state of conciousness and measure and reduce stress thru biofeedback training

strengths of objective data

unable to affect data collection, hence the risk of participant bias is limited

measures can be recorded in real time, allowing researchers to observe physiological responses during set tasks or exposure to stimuli

limitations

exercise can cause changes to physiological measurements (concern if the state of conciousness is being identified)

other factors like heat can affect results

wearable instruments recording physiological responses can cause anxiety undegoing testing, leading to innacurate results

• these act as extraneous variables and confounding if not realised by researchers and controlled

subjective measures

data based on personal opinions and judgements of participants

rating scales

used to quantify abstract concepts, like level of pain

• likert scale: rating scale often used to measure attitudes (5-7 point scale allocating a numerical score to determining whether an attitude is positive or negative

strengths of rating scale

data collected from quantitative subjective measures can be easily statistically analysed than those collected by qualitative methods like interviews

data collected from large sample size using subjective quantitative can be done in a short time (relatively)

can be conducted remotely via mail or online

limitations of rating scale

responses participants can give are limited

not able to give reasons for responses

reading ability is required

way statements are worded and order in which they appear can influence responses

mixed methods

quali/quanti data gathered from participants in same study

using interviews and rating scales together is an example of using mixed method design

strengths

greater understanding of research problem can be provided, as opposed to using quali or quanti alone

can be used to complement each other (e.g use of interview can lead to development of rating scale)

limitations

greater expertise from researchers are rwquired to coll3ect + analyse data

time required for analysis and collection is greater with mixed methods

advantage of using mean

all raw data are accounted for

disadvantage of mean

sensitive to outliers

median

calculated by listing values in numerical order and selecting the value that is located in the middle of the list

advantage of median

not affected by outliers

disadvantage of median

median calculated may not be a number in the original data set if an average of two middle numbers was produced

validity

the extent to which a measurement tool evaluates what it is designed to measure

• mood rating includes statements that allowed for mood to be measured = high validty

• if statements did not allow for mood to be measured, then mood rating scale would be low

reliability

the degree to which a measurement tool produces consistent results

• mood rating scale once a month for three months and results are similar = reliability

test retest reliability

commonly associated with tests such as IQ and questionnaries

can apply to consistency of specific measurements or procedures such as reaction time and heart rate

assesses the extent to which results from an assessment tool are similar when administered to the same participants at two different times

inter-rater reliability

the extent to which diffferent researchers administering the same assessment tool obtain similar results

individual recording test scores or data are raters and for reliability to be measured there need to be two raters to allow for comparison between collected scores or data.

to see if there is inter-rater reliability, must assess correlation between the sets of results and if similar results are calc, then assessment tool has high interrater reliability

how is test retest reliability and intterater reliability assessed

by correlating two sets ofdata and the degree of correlation is statistically measured to produce a correlation coefficient

• deemed reliable when coefficient is at or above +80

• TEST RE-TEST: if not, should redesign test or questionarre, and

• INTER RATER: indicates low level of agreement between raters

internal validity

examines whether a study was designed, conducted and analysed without bias and whether researchers can be sure that changes in dependent were caused by independent and not confouding

external validity

whrther produced results can be generalised to the sample it was taken from

• higher generalisability = greater external validity

e.g conducting experiemnt in environment similar to population it was taken from increases external valiity

connection between internal and external

as you increase control over extraneous variables to strengthen internal validity, ability to generalise findings to broader population and real life settings is limited, thus reducing external validity

generalisability of sample to population

good generalisability = results collected from sample can be applied to population

• new sample should be able to be selected from population, research be replicated and results similar to og sample

generalisability

the extent to which resulst gathered from a sample in research can be applied to other situations

how can study have good generalisability

• sample needs to be representative of population

• used for stratified sampling becays epartiicpants are selected in same proportions in which they appear in population

not knowing if independent variable influenced the dependent variable

use control group to act as baseline for comparision

• indicates whether it is IV affecting DV

extraneous variable

• random allocation when placing into control and experimental groups

• participants are not aware which group theyre in

• eliminate experimenter effect

• monitor controlled variables

• standardised instructions and procedures

• conduct experiment in controlled enviornment

confounding

controll extraneous variables so that they do not turn into confounding