Methods (general) Flashcards
Outline the process of conducting a sign test
1) Remove = scores
2) Add up positive scores and negative scores
3) Whichever is smaller is the calculated value
4) Using the table, find the critical value
5) Check if calculated value ≤ critical value
6) If so, the data is significant
When would you use a directional hypothesis?
When the findings from previous research on a similar topic suggest a particular outcome
When would you use a non-directional hypothesis?
When there is no previous research on the topic, or the findings of previous research on a similar topic have been contradictory
What are the strengths of correlational research?
Doing correlational research is cheap and easy. It can be useful when an carrying out an experiment would be unethical. It can be done before an experimental study to establish patterns.
Evaluate independent groups design
Strengths: not affected by order effects, less susceptible to demand characteristics
Weaknesses: affected by participant variables, requires twice as many participants
Evaluate repeated measures design
Strengths: not affected by participant variables, requires fewer participants
Weaknesses: affected by order effects, susceptible to demand characteristics
Evaluate matched pairs design
Strengths: not affected by order effects, less susceptible to demand characteristics, controls for participant variables
Weaknesses: time-consuming and expensive
What are the limitations of correlational research?
You cannot draw causal conclusions. Sometimes you do not know which co-variable caused the change in the other, and sometimes there is even a hidden variable affecting both variables (the third variable problem).
Evaluate laboratory experiments
Strengths: high control of EVs, good internal validity, good replicability
Weaknesses: artificial tasks, poor external validity, affected by demand characteristics
Evaluate field experiments
Strengths: high mundane realism, good external validity
Weaknesses: low control of EVs, poor replicability, ethical issues with lack of consent
Evaluate natural experiments
Strengths: high external validity, opportunities for research which may otherwise have ethical or practical limitations
Weaknesses: no random allocation, poor replicability, rarely occuring events
Evaluate quasi experiments
Strengths: high control of EVs, good internal validity
Weaknesses: no random allocation, affected by confounding varibales
What are the criteria for the sign test?
Related data (repeated measures design), it is a test of difference, nominal data (categories)
Evaluate random sampling
Strength: free from researcher bias
Weaknesses: time-consuming, may still result in an unrepresentative sample
Evaluate systematic sampling
Strengths: free from researcher bias, fairly representative
Weakness: time-consuming
Evaluate stratified sampling
Strengths: free from researcher bias, representative
Weakness: sub-groups cannot perfectly categorise participants
Evaluate opportunity sampling
Strengths: convenient, time and cost-effective
Weaknesses: researcher bias, unrepresentative
Evaluate volunteer sampling
Strengths: time and cost-effective
Weakness: volunteer bias (attracts a certain ‘profile’ of participants - curious, keen, helpful)
What do we need to know to use the critical values table for a sign test?
Whether the hypothesis is 1 or 2 tailed, the number of pieces of data, and the % of error allowed (usually 5%)
How do we deal with informed consent?
Participants should be issued with a consent letter detailing all relevant information. If the participant is younger than 16, their parents’ consent should be gained too.
How do we deal with deception and protection from harm?
At the end of the study participants should be fully debriefed (made aware of any deception, given the right to withdraw their data, given reassurance about their performance, and offered counselling)
How do we deal with confidentiality?
Participants’ details should either be protected or they should remain anonymous, during briefing and debriefing they should be reminded that their data will be protected
Evaluate naturalistic vs controlled observations
Naturalistic have high external validity, while controlled have good control of EVs and replicability
Evaluate covert vs overt observations
Covert are not affected by participant reactivity increasing validity, while overt are more ethically sound
Evaluate participant vs non-participant observations
Participant have good validity as the researcher gains important insight, while non-participant maintain objectivity
How do we maximise and then measure inter-observer reliability?
The researchers would test their consistency in a pilot study, and then discuss any differences in interpretation. To measure inter-observer reliability the two sets of observations should be correlated (good = a correlation of 1).
Evaluate structured observations
Strengths: recording data is easy and systematic, quantitative data is easy to analyse
Weaknesses: lacks rich observational detail, depends on quality of behavioural categories
Evaluate unstructured observations
Strength: rich observational detail
Weaknesses: qualitative data is difficult to analyse, observer bias
Evaluate event vs time sampling
Event sampling is useful when the target behaviour occurs infrequently, and records all instances of the target behaviour, while time sampling is useful in more complex observations by reducing the number of observations that have to be made.
What are some strengths of questionnaires?
They are cost-effective, can gather large amounts of data, can be completed without the presence of an expert, and produce easy to analyse data that lends itself to statistical analysis
What are some limitations of questionnaires?
They are reliant on truth and memory, and are affected by social desirability bias and response bias
Evaluate structured interviews
They are easy to replicate, but give no room for elaboration or deviation so may miss important information
Evaluate unstructured interviews
They are flexible which means that rich, precise detail can be gathered, but the data produced is difficult to analyse
What five things should be avoided when writing questions for questionnaires and interviews?
Overuse of jargon, emotive language, leading questions, two-part questions, and double negatives
Outline the design of an interview
The interviewer creates a standardised interview schedule, the interview takes place in a quiet room and begins with some neutral questions, the interviewer makes notes or records the answers, and the interviewee is reminded that their answers will remain confidential.
Explain the difference between correlations and experiments
In an experiment the IV is manipulated to measure the DV, but in a correlation there is no such manipulation
What are some strengths and weaknesses of qualitative data?
Strengths: rich detail, insightful, developed thoughts/feelings/opinions, good external validity
Weaknesses: difficult to analyse, subjective interpretation affected by bias
What are some strengths and weaknesses of quantitative data?
Strengths: easy to analyse, lends itself to statistical analysis, objective and less affected by bias
Weaknesses: narrow in scope, poor external validity
What is one strength and one weakness of primary data?
Strength: specifically targets subject of investigation
Weakness: requires time and effort on the part of the researcher
What are some strengths and weaknesses of secondary data?
Strength: inexpensive, easily accessible, little effort required
Weaknesses: variable quality and accuracy, may not precisely fit the subject of investigation
Evaluate the use of the mean as a measure of central tendency
Strengths: sensitive, includes all data, representative
Weakness: easily distorted by extreme data
Evaluate the use of the median as a measure of central tendency
Strengths: easy to identify, not affected by extreme data
Weaknesses: does not include all data
Evaluate the use of the mode as a measure of central tendency
Strengths: easy to identify, appropriate for nominal data
Weaknesses: crude measure, does not include all data
Evaluate the use of the range as a measure of dispersion
Strength: easy to calculate
Weaknesses: easily distorted by extreme data, does not consider all data, unrepresentative
Evaluate the use of the standard deviation as a measure of dispersion
Strengths: precise, includes all data
Weakness: easily distorted by extreme data
In what circumstances may we use a more stringent significance level than 5%?
When research may involve human cost, when there is no possibility of replication, or when the findings have significant social implications.
What are the three main aims of peer reviews?
To allocate research funding, to validate the quality and relevance of research, and to suggest amendments or improvements to research.
Explain how anonymity is an issue in peer reviews
Usually the ‘peer’ doing the reviewing remains anonymous to increase honesty, but a minority of peers have used this anonymity to criticise rival researchers in competition for funding.
Explain how publication bias is an issue for peer reviews
Journal editors are more interested in ‘headline grabbing’ and positive results, so research that doesn’t meet this criteria may be devalued during peer review.
Explain how peer reviews can bury ground breaking research
‘Peers’ are likely to criticise research which contradicts mainstream theories or their own views, slowing down the publication of new and innovative research