research methods Flashcards
what is an experiment?
A testable situation that determines a cause-effect relationship between variables under controlled conditions
what is an independent variable?
the variable that is systematically manipulated by the experimenter in order to measure its effect on the DV.
what is the dependent variable?
the variable that is used to observe and measure the effects of the IV
what is an operationalised variable?
turning the IV and DV into procedures so that it can be manipulated or measured in an experiment
what is a hypothesis?
a statement or testable prediction about the direction of interaction between the IV and DV and the population from which the sample is drawn.
what is a threory?
general explanation of a set of observations or findings of behaviour and/or mental processes that seem to be related
what is a model?
representation of how some behaviours and/or mental processes could, should or do occur
define population
everyone of research interest from which the sample is selected
what is a sample?
the group of people taken from the population to be studied
what is sampling? what is the goal of sampling?
process of selecting participants for a research study to obtain a representative sample of the population from which it was selected
what is stratified sampling?
divides a population into strata (subgroups) based on particular differences in characteristics and then participants are selected from each stratum in the same proportions as they exist in the population
evaluate stratified sampling
adv:
- more representative and unbiased
- allows researcher to sample specific strata for comparative purposes
dis:
- time consuming; requires list of target population for information about population characteristics to be available
what is random sampling?
every individual in the population has an equal chance of being selected for the sample
evaluate random sampling
adv:
- more representative
- generalisations are more likely to have external validity
dis:
- still a chance the sample is unrepresentative
- requires complete list of the target population; time consuming if list is unavailable
what is convenience sampling?
selecting participants who are readily and easily available
evaluate convenience sampling
adv:
- quick, easy, inexpensive
- can be used as a test procedure or gain an indication of possible responses before the actual study
dis:
- biased sample; only those available at the time and location of the study have a chance of being included in the sample
- data obtained can be misleading= results cannot be generalised
what is a representative sample and why is a bigger sample more representative?
- sample that is approximately the same as the population in every important participant variable
- as sample size increases, the characteristics of the sample more closely reflect the characteristics of the population from which the sample was drawn
define allocation
the process of assigning participants to either the experimental group(s) or the control group
what is the purpose of the control group?
not exposed to the IV to provide a baseline of comparison for the experimental group to determine if the IV has had any effect on the DV
what is random allocation?
every participant in the sample has an equal chance of being in the experimental or control group
what is the advantage of random allocation?
- Minimises the effects of individual differences that may occur between participants
- more likely to spread the participant characteristics evenly across the two groups
what are the types of experimental designs?
- independent groups
- matched participants
- repeated measures
what is independent groups design?
different participants are used in each group of the experiment and are exposed to different conditions
evaluate independent groups design
adv:
- if random allocation and sampling is used= helps increase the likelihood of an even spread of participant characteristics between conditions
- no order effects occur as participants only complete one condition
- experiment can usually be completed on one occasion= minimal participant dropout
dis:
- least effective at controlling individual differences between conditions
- often need a larger number of participants to help the spread of participant characteristics
what is matched participants?
each participant is paired with a participant in the other condition according to any important variables relevant to the experiment and then each member of a pair is allocated to either experimental or control group
evaluate matched participant design
adv:
- participant variables are minimised as for every participant there is a similar participant in the other condition
- no order effects as participants only complete one condition of the experiment
dis:
- pairing participants could be time-consuming as pre-testing might be needed
- cannot match participants perfectly on every relevant variable
- loss one participant means loss of the whole pair
what is repeated measures design?
one groups of participants undertake both conditions of the experiment
evaluate repeated measures design
adv:
- participant variables are eliminated as they are kept constant between conditions
- usually fewer participants are required
dis:
- order effects can occur when performing the 2nd condition eg. practice, boredom, fatigue
- participant drop-out rate tends to be higher
what are extraneous variables? why are they not cool and quirky?
- Any variable other than the IV that can cause a change in the DV and therefore affect the results of the experiment in an unwanted way
- makes it difficult for researchers to conclude that any change in the DV was caused solely by the IV and not some other variable
what are confounding variables?
A variable that systematically varies with the IV that has had an unwanted effect on the DV, making it impossible to determine which of the variables has produced the change in the DV.
what are the differences between evs and cvs?
ev:
- May or may not have an effect
Reduces reliability of results
Effect is usually small
cv:
- definitely altered the DV
- Has a similar, although often a greater effect than the IV
- cannot draw a conclusion or make generalisations
- Reduces validity
what are 6 types of EVs and CVs?
EVs:
- individual participant differences
- experimenter variables
- situational variables
- non-standardised procedures and instructions
CVs:
- order effects
- placebo effect
what are individual participant differences? how does it affect the dv?
- The unique personal characteristics, abilities and backgrounds that participants bring with them to the experiment
eg. sex, intelligence, personality characteristics, motivation, emotional state, mood, problem-solving ability, self-esteem, health, cultural background - can affect the way participants responds and their performance in an experiment
what are situational variables? how do they affect the dv?
- any variable associated with the experimental situation itself that may affect the results of an experiment
eg. Background noise, time of the day, testing venues, testing conditions, air temperature - participants in different groups may not be tested under the same conditions and in the same situation= experimenter cannot confidently conclude that any change in the DV is the result of the IV.
what are experimenter effects and why do they affect the dv?
- variables associated with behaviour, expectations, biases, characteristics of the experimenter during the experiment which may result in the participants behaving in a manner different from how they would normally behave= altering the results
eg. The experimenter being tired, their expectations of the outcome of the research, personal issues, health, their gender, how they dress and their attractiveness
what are use of non-standardised instructions and procedures?
Not the same:
Procedures= everything the researcher does in conducting their study
Instructions= what the participants are told
what are order effects?
When performance of one task influences the next task because of the specific order in which the conditions are presented rather than the IV
- cannot determine whether order of condition completed or IV caused the change
eg. Practice effects: repeating a task – can get better or worse in the second condition
-> Boredom, fatigue
Carry-over effects: influences a task has ‘carries over’ to the next task
-> Medication, mood
what are placebo effects?
Any change in a participants behaviour/improvement in symptoms triggered by the belief they are receiving some experimental treatment= respond in accordance with that belief➝ expectation of feeling better
- cannot determine whether the expectations or the IV caused the change
what is a placebo?
- fake/inactive treatment/substance that appears real/resembles the real treatment
how do you control for individual participant differences?
- the type of experimental desgin
- use of random sampling and allocation
how do you control for non-standardised procedures and instructions?
standardised procedures: techniques used to measure participants responses are identical and all participants are treated in the same way, appropriate to the condition they have been allocated to
and instructions: instructions given to all participants for each condition have been predetermined and are identical
how do you control for order effects?
Counterbalancing
- systematically changing the order of the treatments for participants, allowing order effects to occur more evenly across both conditions to control for the unwanted effects
how does counterbalancing work?
- Randomly allocate the sample into two groups
Group 1: Control condition > experimental condition
Group 2: Experimental condition > control condition - The results for all participants are then combined across the entire experiment to cancel out order effects
how do you control for experimenter effects?
Double-blind procedure
- where neither the participants nor the experimenter knows whether the participants are in the experimental or the control group
how do you control for placebo effects?
Use a placebo
- All participants now receive the ‘same’ treatment
= Any differences in the DV can then be attributed to the IV because both groups form the same expectation
Single-blind procedure: Where the participants don’t know if they are in the experimental or control group= prevent forming expectations & therefore changing behaviour
what are the 5 types of research methods?
- experimental research
- Cross-sectional studies
- Case studies
- Observational studies
- Self-reports:
-> Questionnaires
-> Interviews
-> Rating scales
evaluate experimental research
adv:
- IV can be manipulated= possible to test whether there is a cause and effect relationship between the IV and the DV
- easy to repeat experiment
- more controlled environment
dis:
- artificial environment= participant behaviour may change to the point it cannot be generalised to situations outside a lab
what are cross-sectional studies?
- Selects and compares groups of participants on one or more variables at a single point in time
- uses and measures existing, naturally formed groups and characteristics in sample/pop= not a true experiment because people cannot be randomly allocated to groups
eg. Age differences in short-term memory – a researcher could select groups of 10, 20, 30, 40, 50, 60 & 70 year olds to test then compare results
evaluate cross-sectional studies
adv:
- can study change over time without needing to wait for long periods
- simpler, quicker, less expensive
- can conduct research on topics that would be unethical in experimental situation
dis:
- cause-effect relationship cannot be tested
- individual participant differences
- cohort effect: generation gaps that impact thoughts, feelings and behaviours
what are case studies?
intensive, in-depth investigation of some behaviour or event of interest in an individual group, organisation or situation
- used when large numbers of participants are not available -> < 3
- no manipulation of IV= not an experiment
eg. H.M. (amnesia due to temporal lobe damage)
evaluate case studies
adv:
- provide detailed information
- no extraneous variables to consider
- provides details on unusual cases
- avoids artificiality of labs -> actual experiences
dis:
- low reliability= cannot be replicated
- often subjective= difficult to collate and analyse data
- small sample size= difficult to generalise as results may not reflect typical ways of thinking, feeling and behaving
- can be biased -> participants may withhold info + experimenter effect can occur
what is observational studies?
collecting data by carefully watching and recording behaviour as it occurs
- not considered a true experiment unless random allocation is used
eg. observing children to see if there are any gender differences in aggressive behaviour
what is naturalistic and contrived observation?
- researcher records naturally occurring behaviour in a non-influential manner
- a researcher creates an artificial environment for observation to occur
what is participant and non-participant observation?
- researcher actively engages in study being observed
- researchers observe from the ‘sidelines’
evaluate observational studies
adv:
- record behaviours in their natural setting= not influenced by artificial environment
- can see long term changes
dis:
- can be time-consuming when waiting for behaviour to occur naturally
- ethical issues of informed consent
- difficult to determine causes of behaviour,
- researcher bias – unconsciously distorts what they see to match their research criteria
what are self-reports? what are the three types of self-reports?
participant’s written or spoken responses to questions, statements or instructions presented by the researcher
- questionnaires
- interviews
- rating scale
what are questionnaires? evaluate
written set of questions in a structured format
- can be used with large number of participants
- do not require supervision= efficient
what are interviews?
face-to-face meetings with participants that can be structured (have set questions) or unstructured (use a variety of questions and discussion)
what are rating scales?
fixed-response questions/statements where participants rank each item by selecting from a number of choices
evaluate self-reports
adv:
- efficient way to collect data from a large number of people in a short period of time (not interviews)
- appropriate to use for sensitive topics due to anonymity
dis:
- data can be difficult to summarise, collate and analyse
- language dependent: people with weak literacy skills are disadvantaged
- give false/misleading answers due to wanting to create a favourable impression
- non-response bias: people who choose to respond may be different to people who don’t respond
- can’t accurately recall info
what is primary data?
Collected directly by the researcher for their own purpose
eg. Participant answers on a survey
what is secondary data?
Collected by someone other than the original user for their own purpose and will not be used for the first time
eg. statistics from the ABS
what is qualitative data?
information about the or characteristics of what is being studied
eg. Descriptions, words, meaning, pictures
what is quantitative data?
Numerical information on the amount of what is being studied
eg. Units of measurement or numbers (raw scores, percentages, means, standard deviations)
evaluate primary data
adv:
- offers tailored information sought by the researcher to test their hypothesis on their chosen topic.
dis:
- collection can be time-consuming to process
evaluate secondary data
adv:
- tends to be readily available and can usually be accessed in less time
dis:
- there can be uncertainty about its quality because it was collected for another purpose
evaluate qualitative data
adv:
- covers issues in great depth and detail through interaction with participants.
- it is flexible and creates openness during research
dis:
- can be subjected to the biases of the researcher
- Interpretation and analysis can be time-consuming.
evaluate quantitative data
adv:
- allow researchers to easily summarise vast sources of data.
- More precise and detailed analysis can occur
dis:
- can provide less detail on motivation, attitudes and behaviour of participants
what is objective data?
info that is observable, measurable & free from personal bias of researcher
what is subjective data?
info based on personal opinion, interpretations, and judgement determined by participants
- usually not verifiable
what are descriptive statistics?
- used to organise, summarise, analyse and describe results
- Often displayed in tables and graphs which organise data to display how often each score occurs
how do you present a table?
Table 1
Title in Italics
- written at the top of table
how do you present a graph?
- written at the bottom of graph
Figure 1 in italics. Title not in italics
what are the differences between the 3 graphs?
bar graph:
- seperate bars
- x-axis: category, y-axis: frequency
histograms:
- touch= shows continuous data
line graphs:
- shows relationship between 2 variables
what are percentages?
expresses a number as a proportion out of 100
what is measure of central tendency?
calculation that shows how typical scores, or a majority of scores fall in a data set
eg. mean, median, mode
what is the mean? why is it useful? what are the limitations?
- Arithmetical average of all individual scores in a set
- if scores cluster closely around a central score= mean is fairly accurate indicator of the typical score
- can be influenced by outliers= not representative or accurate of typical scores
what is measure of variation?
indicate how widely scores are distributed or spread around a central score
eg. standard deviation, range
what is standard deviation?
summarises how far scores within a set are spread from the mean
- Low SD: little variation in the scores from the mean
- score in data set will be similar
- High SD: greater variation in the scores from the mean
why is standard deviation useful?
- provides a point of comparison between the means and the spread of scores in two (or more) data sets
what are inferential statistics?
used to make judgements about what the data actually means
- allow us to infer a cause-effect relationship
- allows experimenter to draw conclusions about their results
- limitation to conclusions if inferential statistics are not calculated
what are statistical significance?
significance of the difference between the results of the experimental and control group and whether the difference is because of the IV or chance
- determines whether or not a results support the hypothesis
what are p-values and how are they calculated?
level o f probability that the difference in results are due to chance
- determines statistical significance
- t-test compares the distribution of results of two conditions
- An acceptable p-value is established before the research is conducted
what does p < 0.05 mean?
the probability that the results are due to chance (and not the effects of the IV)
is less than
5% or 5 in 100 repetitions of the study
- If the likelihood of the difference occurring by chance is extremely low, then it can be said that the difference is “statistically significant”
what is the advantage of using statistical significance?
- more efficient way to determine reliability= tests the extent of chance factors affecting the results
- rather than repeating the study with the same participants= time-consuming & impractical
what is a conclusion? what are the considerations to make before drawing a conclusion?
A decision about what the results from the study mean (in the sample)
- Results of the experiment
- Reliability and validity
- Presence of confounding variables
conclusion formula
In this study, the hypothesis was supported/rejected. It can / cannot be concluded that the experimental group (state evidence/result) has performed high/lower/better/worse than the control group (state evidence/results)
E.g. In this study, the results supported the hypothesis. It can be concluded that in this study adolescents who were sleep-deprived (average number of errors being 7) have decreased cognitive abilities than those who were not sleep-deprived (average number of errors being 4)
what do you do if you decide to draw a conclusion?
- A conclusion is limited when based on descriptive statistics - inferential statistics (p-values) must be calculated to determine the probability results were due to chance factors
- Any EVs present would need to be controlled in future replications of the study (explain how)
- Repetitions of the study could be conducted to determine whether the results can be generalised to the target population / potential other populations (and also further determine the reliability of results)
what do you do if you can’t draw a conclusion?
explain:
- Relevant EVs/CVs that may have affected the results (reducing the validity of the study) and make suggestions to minimise their impact in the future (thus improving validity)
- If the study was to be repeated, this will determine whether the result obtained was a ‘one-off’ or if the IV consistently does not affect the DV (reliability of results)
what is a generalisation? what conditions must be met to make. a generalisation?
a decision about how widely the findings of the research can be applied to members of the research population
- The sampling method and the sample (i.e. the sample must be a suitable size, representative of the population and convenience sampling should not be used)
- The experimental procedures
- Presence of confounding variables/validity
- statistical significance
generalisation formula
A generalisation can / cannot be made to the insert population of research interest relevant to study because refer to the sampling technique and explain why the sample is likely / not likely to be representative of the population
what is reliability?
The extent to which results obtained from research are consistent or stable.
high reliability= results are highly consistent over multiple replications of the study
what is validity?
The extent to which the procedures or methods in the study measured what they were supposed to measure
high validity= accurately measured the effect of the IV on the DV
what is internal validity? how can it be improved?
the extent to which the results in a study are due to the IV and not other factors
- Using an appropriate research design
- Controlling extraneous variables
what is external validity? how can it be improved?
the extent to which the results of a study can be generalised to the population from which the sample was drawn or other people/groups
- Conducting experiments in settings natural to the topic of interest
- Ensuring the sample is representative of the population
why can you have a measure that is reliable but not valid?
You can obtain consistent results without measuring what you are supposed to measure
why is a valid measure always reliable?
if you are accurately measuring what you are supposed to be measuring (e.g. weight) the results must be consistent
what are ethics?
standards that guide our conduct
- primary purpose is to protect the rights and welfare of human participants
- Secondary purpose is to ensure the research done is to the benefit of human society
what are the 4 basic ethical principles?
- Integrity: committing to the search for knowledge and the ethical conduct of research
- Respect for persons: regarding the welfare and rights of participants
- Beneficence: maximise possible benefits and minimise discomfort or harm to participants
- justice: fair distribution of benefits and burdens within a population of research interest
what is deception and debriefing?
deception: when participants are deliberately mislead or not fully informed about the aim or nature of the research.
- Only used when knowledge of the true nature of the study will influence the behaviour and responses of the participants
debriefing: After the study has concluded, the purpose, results and conclusions of the study must be clarified and the wellbeing of the participants must be checked
what is informed consent and voluntary participation?
Individuals must have a full understanding of the purpose, methods, demands, risks and potential benefits and what their participation entails and must voluntarily agree to all aspects of the study usually by written consent
voluntary participation: Participants willingly choose to be involved in a study and may not be pressured to take part in the experiment, nor experience negative consequences if they do not participate
what is confidentiality and withdrawal rights?
confidentiality: The researcher must not use or disclose private information about participants for any other purpose other than for which it was given to them
withdrawal rights: participants are free to decline their participation or pull out of a study at any time, even after the study has concluded and without reason. This is explained before a study commences.
