research methods 2

qualitative data

involves find out what people think and how they feel in detail, produces descriptive data

where is qualitative data used

used in unstructured interviews, case studies

strength of qualitative data

more detailed, offers explanations, external validity is higher

weakness of qualitative data

harder to analyse, internal validity is low due to potential bias interpretation of data and lack of objectivity

quantitative data

involves measuring something, to produce numerical data (numbers)

where is quantitative data used

used in experiments, research

strength of quantitative data

easier to analyse and easy to interpret findings, higher internal validity due to lack of subjectivity

weakness of quantitative data

less detailed, doesn't offer explanations, external validity is lower

primary data

information collected directly from participants by researchers for their own research aim

strength of primary data

authentic data from the p's for the purpose of the investigation, extraneous variables are controlled

weakness of primary data

requires time and effort - designing and collecting questionaries takes time and expense

secondary data

information that already exists somewhere, having been collected for another purpose

strength of secondary data

inexpensive - the desired info may already exist, so it minimises effort and the researcher having to find out his own data

weakness of secondary data

quality may be out of data - info may be out dated or incomplete, so it challenges the validity of any conclusion

meta analysis

a form of secondary data collection whereby a researcher gathers multiple research papers on the same or similar topics

strength of meta analysis

usually a large sample size due to multiple research studies increases generalisability, and validity

weakness of meta analysis

Lack of control over integrity of original data. Original data should be interpreted with caution.

descriptive statistics

allow us to describe and summarise quantitative data

measures of central tendency

mean, median, mode

strength of mean

uses all scores so most powerful and sensitive measure

weakness of mean

can be distorted by anomalies

strength of median

not affected by anomalies, easier to calculate than mean

weakness of median

only takes into account middle values

strength of mode

unaffected by anomalies, easier to calculate than mean

weakness of mode

not useful in small sets of data, or when there are too many modes, doesn't take into account other scores

measures of dispersion

information about how spread out the scores are (range and standard deviation)

strength of range

strength; easy to work out (subtract smallest number from the biggest), takes full account of extreme scores

weakness of range:

distorted by anomalies, doesn't take it all scores

standard deviation

a more sophisticated measure of variability, takes into account all scores and their difference from the mean value

strengths of standard deviation

takes into account all scores, a more sensitive measure of spread

weakness of standard deviation

less meaningful as data is not normally distributed

graphs and tables

useful for summarising data as they allow psychologists to see patterns in the data very easily

bar chart

presents non continuous nominal data; when variable falls into categories, bars should be kept separate

histograms

shows continuous data and useful for showing the distribution of scores, bars should touch each other

pie chart

a chart that shows the relationship of a part to a whole

scattergrams

solely for displaying correlations

results tables

help to summarise the main findings of data, which data differ from raw data tables that only contain unprocessed scores

normal distribution

most people will gain a score that centres on the mean. median, mean and mode occur at the same place; at the peak of the curve

negatively skewed distribution

outliers significantly affect the mean than other averages, will contain significantly more high scores than low scores and can be classed as having a ceiling effect, mean will be lower than the mode and median

positively skewed distribution

contains more low scores than high scores, most of the scores falling below the mean, will have a floor effect, mean will be higher than the mode and median

levels of measurement

nominal, ordinal, interval

nominal data

a frequency count for distinct categories where something can only belong to one category; eg - no. of people who smoke

ordinal data

numbers can be placed in ascending or descending rank order; eg - coming 1st or 2nd or 3rd on a test, we do not know the true intervals of each position

interval data

measurements are taken from a scale where each unti is the same size the gap between each unit is fixed and equal. eg; temperature, height

chi squared test

nominal data and an experiment using independent measures design

sign test

nominal data, repeated measures/ matched pairs designs experiment

mann whitney u

ordinal data, independent measures design experiment

wilcoxon t

ordinal, repeated measures/matched pairs designs experiment

spearman's rho

ordinal data, correlation

unrelated t test

interval data, independent measures design experiment

related t test

interval data, independent measures design experiment

pearson's r

interval data, correlation

type 1 errors

wrongfully accept the experimental (alternative) hypothesis, believing there is a difference or relationship, when actually no such relationship exists

type 2 errors

wrongfully accepting the null hypothesis, believing there is no difference between conditions or no relationship when in fact, a relationship does exist

peer review

the process where other experts in the field anonymously evaluate a research study before it is published to ensure its quality, validity, and originality

peer review process

scientists study something, write about their results, a journal editor receives an article and sends it out for peer reviews, peer reiviwers read the article and provide feedback to the editor, if an article finally meets editorial and peer standards, it is published in a journal

advantages of peer review

allocates funding to worthy institutions and allows for scientific fraud to be identified before publication

disadvantages of peer review

biased in their recommendations to the journal editor so valid research being rejected, there's no anonymity, so if experts have conflicting research, they may not approve research to further their own work, gender bias

economical impact of effective psycho therapies

financial burden on the NHS, eg CBT is effective especially after 15-20 sesssions with a trained specialist which can be a burden ont he NHS

economical impact of eyewitness testimony research

can lead to wrong conviction based on inaccurate memory, increasing the time and costs associated with solving causes

economical impacts of attachment

research about fathers being able to be a nurturing attachment figure, can decrease gender pay gap, laws about maternity leave and paternity pay may change, as fathers are being shown as effective primary caregivers

seven features of science

objective, empiricism, replicability, falsifiability, paradigm, popper's hypothetico-deductive model, kuhn's paradigm and paradigm shift

objective

research is unbiased

empiricism

evidence should be gathered through direct observation

replicability

ability to repeat research and achieve the same findings from original study

falsifiability

findings should be possible to refute

paradigm

an agreed subject matter and set of procedures

why is inferential statistics used

used to determine whether results are statistically significant and not due to chance, allowing researchers to accept or reject the null hypothesis.