Research Methods Flashcards
correlation
a measure of how strongly 2 or more variables are related to each other
is correlation more ethical
yes because nothing is manipulated
positive correlation
as one variable increases, so does the other
negative correlation
as one variable increases the other decreases
correlation coefficient
measures the strength of relationship between -1 and +1
2 types of hypothesis
null
alternate (one or two-tailed)
two-tailed hypothesis
“there is a correlation between”
either positive or negative correlation
one-tailed hypothesis
“there is a positive correlation” or “there is a negative correlation”
null hypothesis
“there is no correlation between”
nominal data
data put into named categories
ordinal data
data put into rank order from lowest to highest
interval data
data is taken using the same unit of measure throughout the range
discrete data
data that can only take a certain individual value
continuous data
data that can have any number value in a certain range
assumptions of parametric tests
populations drawn from should be normally distributed, variances of populations should be equal, at least interval/ratio data, no extreme scores
why non-parametric tests are used
when assumptions of parametric tests can’t be fulfilled, when distributions are non-normal
independent variable
the factor you change
dependent variable
the result when affected by the independent variable
experiments designs
independent groups, repeated measures, matched groups, quasi
independent groups
when there are groups of different participants being tested by the same test
repeated measures
the groups of participants are the same but given different tests
matched groups
the different groups are matched for intelligence/ability/skill etc.
quasi
an experiment with independent variables that are not open to manipulation therefore, have naturally occurring groups
if the standard deviation is large
the values are widely distributed and scores can occur a long way from the mean
symmetrical distribution
mean = mode = median
positive skew
histogram leans to left
mode < median < mean
negative skew
histogram leans to right
mean < median < mode
central tendency and spread most suitable for skewed distributions
median as mean would be affected by skewed value, IQR for spread
central tendency and spread most suitable for symmetrical distributions
mean, standard deviation for spread
most suitable graph to show how data is distributed
histogram
symmetrical distributions on a boxplot
median in center, whiskers equal length
UQ - median = median - LQ
negative skew on boxplot
median closer to UQ
UQ - median < median - LQ
type 1 error
when you accept hypothesis but null hypothesis is correct
type 2 error
when you accept null hypothesis when hypothesis is correct
how to minimise type 1 and 2 errors
ensuring experiments are controlled, reliable and valid
descriptive statistics
tables, graphs, mode/median/mean
measures of dispersion
variance: average of the squared differences from the mean
range: spread of scores
standard deviation: a measure of how spread out the scores are (square root of variance)
measures of central tendency
mean, mode, median
standard deviation
measures the spread of the set of numbers (average of each number from mean)
the bigger the standard deviation
the bigger the spread of numbers
sample techniques
opportunity, random, self-select/volunteer, snowball, clinical
purpose of binominal sign test
difference between 2 conditions when data falls into 2 categories
alternate hypothesis
predicts there will be a relationship between the variables and the results are significant
target population
the group of people the researcher is interested in and draws the sample from
random sampling
each member of the target population has an equal chance of being chose so no bias and likely to be representative but no control over who is selected
snowball sampling
relies on initial participants recruiting others to generate additional participants, sample is unlikely representative but easy if specific sample features required
opportunity sampling
selecting people most easily available at the time of study, unlikely representative but quick and easy
self-selected sampling
people asked to volunteer in the study, unlikely representative but those are usually willing and cooperative so perform best to ability
what do you put into the abstract
brief statement of the topic investigated, design used, participants, stimulus materials, apparatus, brief results/analysis/conclusions
what do you put into the introduction
review of background material, outline of the topics being investigated, aim and hypotheses
what do you put into the discussion
interpretation and meanings of results, limitations of the study, short statement of results, account/explanation of findings, how it has benefited the community and improved understanding
what do you put into a harvard reference
surname, first initials, year published, title, book, pages
what do you put into the appendices
raw data, statistical calculations, stimulus material, standardised instructions, copies of questionnaires etc
what is the purpose of a peer review
to validate new knowledge and ensure integrity, researchers comment on work, follow it up, produce further research
