Research Methods & Analysis
What are extraneous variables and confounds?
Variables that should be kept constant, because they can influence the the results of the study. If extraneous variables are not controlled, confounding variables will cause a change in the DV instead of the experimental conditions of the IV.
What is the control group?
The group that does not recieve the treatement that is expected to cause a change in the DV - can be used to compare with other experimental conditions.
What is the placebo group?
recieves treatement. but participant is unaware of what group/experimental condition they are in - in order to view the real affects of the treatement.
What is a hypothetical construct
A non-measurable and non-observable variables - can not be measured using scales and is usually inferred from behaviour. (e.g. intelligence)
Define 'operationalizing'
The act of converting a hypothetical construct into a variable with a measurable scale
Define 'operational definition'
The set of activities required to measure parts of a hypothetical construct.
Validity vs Reliability
Validity: ACCURACY OF MEASURE - the extent to which the results really do represent what they say they will/ measures actually asseses what it is supposed to.
Reliability: CONSISTENCY OF MEASURE - is the measurement tool consistent, can the results be reproduced under the same conditions?
List the 'levels of measurement' to measure variables
Nominal, Ordinal, Interval, Ratio
Nominal vs Ordinary scale
Nominal and orindary: both categorical
Nominal is not ranked and usually used for classification (e.g.: marital status - widowed, divorced, single)
Ordinary is ranked and follows a particular order (e.g. good, very good etc.)
Interval vs Ratio scale
Both numerical/qualitative
Interval: interval between data points is identical but true zero does not exist. (e.g. temperature [can not investigate an experimental condition where there is 'no temperature')
Ratio: "" true zero exists. (e.g. number of times been married)
Define: 'experimenter expectancy effects'
Observer/experimenter may very subtly and subconciously convey to the participants the results they are expecting, thereby leading them down a particular path.
Define: participant/subject expectancy effects
When research participants alter their behaviour or reponse during an investigation because they have an idea of what the experimenter is expecting to see in the results.
How are expectancy effects dealt with in research?
Standardized procedures: identical instructions and procedures for all participants in a spceific condition.
Single blind: participants are unaware of the condition they are in.
Double blind: participants and oberservers are unaware of the conditions they are in. - prevents experiments' expectations and particpants' beliefs from interferring with and influencing the outcome of the study.
Define: 'sampling frame'
A list or database of from which a researcher draws a sample of individuals
What is a sampling bias
If the results from a sample are consistently and repeatedtly different from the truth about the wider population
Convenience vs voluntary response sampling
Both non-probability
Convenience:researchers select the most readily available people/participants.
Voluntary response: Individuals choose to participate in a research study on their own - usually because they have an interest in the topic being studied.
Define: systematic sampling
Sampling method whereby every nth participant is selected from the population, where n is any number.
Stratified vs cluster sampling
Stratified: members of a population are divided in subgroups called strata, members from each stratum are selected.
Cluster: members of a population are divided into subgroups. A few subgroups are selected, then every individual within that subgroup is selected.
Explain why representation of a wider population is important and stressed within research?
The goal of many studies and investigations is to draw a conclusion about certain behaviours, actions and phenomenons that apply to a wider population. If the sample is bias and not representative, it becomes challenging to generalize research findings.
Experimental Designs: independant samples vs repeated measures
Independant samples: independant and unrelated samples of participants in each experimental condition.
Repeated measures: the same participants in a sample complete all the experimental conditions.
Pros and cons of independant samples and repeated measures
Independany samples-
Pros: avoids order effects, carryover effects, attrition rates are lower, lower chance of participants finding out the aim of the study.
Cons: requires more participants and resources, more participants makes it harder to control extraneous variables [increasing the likelihood of the presence of confounds].
Repeated measures-
Pros: Practical in terms of recruitement, variations in individuals are more controlled.
Cons: carryover effects, order effects, attrition and risk of participants finding out the aim of the study, thereby increasing the risk of participant expecancy effect.
Define 'order' effects
If a participant carries out several experimental conditions, they might improve with practice or deteriote due to fatigue. So differences in the results seen will be due to one of these factors rather than manipulation of the IV = confounding factor.
How to deal with order effects
Counterbalancing and complex counterbalancing: systematically changing the order in which experimental conditions or treatements are being presented to the participant.
Elapsed time: control and monitor the duration of time between conditions.
Quasi experiments are present when _____
organismic factors/variables are your IV [such as age, ethnic origin],
so these variables can not be manipulated and we can not randomly assign people to these groups.
Pros [3] and cons [2] of internet-mediated research
PROS: researchers can reach a more diverse and geographically dispersed sample of participants, more cost effective as it can reduces or eliminates the need for physical equipement or materials, digital data is easy to organize which reduces the risk of data loss
CONS: wifi connectivity issues and compataibility issues can influence the participants' experience = research environement isnt controlled. Can not be used if live human interactions or physical materials are required.
What is 'Field study/naturalistic design'?
A non-experimental study conducted outside the lab and often in the natural environement. Observational methods are used to collect data, such as oberving behaviours and interactions, in a non-intrusive manner.
There is a tradeoff between a lack of control and a more realistic setting - so variables are not manipulated, instead, researchers observe how these variables naturally manifest in the world.
= yields qualitative rich data
What is the difference between experimental and non-experimental research design
Experimental: aim is to establish a cause and effect relationship between variables, by manipulating the IV and observing its influences on the DV.
Non-experimental: aims to observe and analyse naturally occuring and existing relationships and correlations, without intervention through the manipulation of experimental conditions.