Correlations Flashcards
correlation
correlation is used to analyse the association between two continuous variables (co-variables)
three types…
• positive correlation
• negative correlation
• zero correlation
what are co-variables?
the two measured variables in a correlational analysis
the variables must be continuous, meaning that they can take on any value within a certain range and can be arranged in an order
positive correlation
two variables increase simultaneously
negative correlation
as one variables increases, the other decreases
zero correlation
there is no relation or link between the two variables
linear correlation
when all the values form a straight line from either bottom left to top right or top left to bottom right
curvilinear correlation
the relationship is not linear, which causes a curved line on the graph
for example, stress and performance do not have a linear relationship. performance on many tasks is low when stress is too high or too low, performance is best when stress is moderate
correlational hypothesis
a study that uses a correlation analysis needs a correlation hypothesis which states the expected association between the co-variables
possible hypotheses might be…
- positive correlation, directional hypothesis — the two variables are positively correlated
- negative correlation, directional hypothesis — as one variable increases, the other decreases
- positive/negative correlation, nondirectional hypothesis — the variables are correlated
- zero correlation, null hypothesis — the two variables are not correlated
scattergrams
a correlation is illustrated using a scattergram, which is a graphical representation of the association between two sets of results
for each individual, 2 scores are obtained which are used to plot one dot on the graph
the covariables determine the x and y position of the dot
the scattered dots indicate the degree of correlation between co-variables
correlation coefficient
researchers use a statistical test to calculate the correlation coefficient which is a measure of the extent of correlation that exists between co-variables
a correlation coefficient is a number between +1 (a perfect positive correlation) and -1 (a perfect negative correlation)
coefficients are written with signs which show whether it is a positive or negative correlation
the coefficient tells us how closely the co-variables are related and how strong the correlation is
researchers must find out if the correlation coefficient is significant — in order to do this they use tables of significance which tell them how strong the coefficient needs to be in order for the correlation to count as significant/meaningful and the hypothesis under test to be accepted
difference between correlations and experiments
in an EXPERIMENT, the investigator deliberately changes the independent variable in order to observe the effect on the dependent variable
without this deliberate change no causal conclusions can be drawn
in a CORRELATION, the variables are simply measured. no deliberate changes are made therefore no conclusion can be made about one co-variable causing the other
for example, a study may show that there is a positive correlation between attendance at school and academic achievement but a researcher could not conclude that the level of attendance caused the better achievements, they’re simply related
limitations of correlations
the variables are simply measured. no deliberate changes are made therefore no conclusion can be made about one co-variable causing the other. for example, a study may show that there is a positive correlation between attendance at school and academic achievement but a researcher could not conclude that the level of attendance caused the better achievements, they’re simply related
correlational research may lead people to jump to causal conclusions, this is a problem because such misinterpretation of correlation may mean people design programs for improvement based on false premises. for example, a headteacher may believe that higher attendance causes better academic achievement and expect improving attendance to improve exam results, such a connection is not justified by a correlation
variables are not controlled which means possible intervening variables can explain why the covariables being studied are linked. for example, it might be that students who do not attend are the ones that dislike school and this dislike also impacts on exam performance so it’s not necessarily lack of attendance that causes poor exam results
strengths of correlations
can be used to investigate trends in data and if a correlation is significant then further investigation is justified
procedures in a correlation can usually be easily repeated, meaning that findings can be confirmed
allows rea search to be conducted and a relationship to be established between two variables where it would otherwise be unethical or impractical to manipulate these variables
