Please enable JavaScript.
Coggle requires JavaScript to display documents.
Research Methods (Data handling and analysis) PT5 - Coggle Diagram
Research Methods (Data handling and analysis) PT5
Quantitative data (numerical) quantitative date (more in depth)
Descriptive statistics
ways to display quantitive data
measure to central tendency - these will give average of the data (mean, median, mode)
mean - when there arena extreme values
median - used alternatively if there are extreme values
mode - when data is all very similar
can have lots of modes but only one mean and median
Primary and secondary data
meta-analysis - physiologists use other researchers data all from the same concept - drawing wider conclusions on the data
meta-analysis - used frequently for drugs , see mass about of inconsistency can draw results that it isn't working
Measures of dispersion
how varied your data is
range - use the range if I have used the median or mode
standard deviation - use if you have used the mean - shows you how far away each participant is from the mean
if measure of dispersion between each participant is really high you cannot conclude accurate results
Postive, negative and zero correlations
positive - both variables increase at the same time
negative - one variable increases, the other decreases
zero - the variables have no relationship with each other
Presentation of data
present mean, mode or median in graphical display
might be asked to draw scattergrams, bar charts or histograms
scattergrams - if you are looking for a correlation (relationship)
bar chart - when you have categories that are distinctively different from eachother
histograms - when data is continuous
Distribution graphs
measure of central tendency (giving average of data)
3 types of skews
postive skews - when your mean is significantly higher than your mean and mode (majority really small data and one or two having a really high score, pushing everything out of sync)
symmetrical skew - very similar (no anomalies, equal spreading of data)
negative skew - mode is significantly higher than the mean (majority of the scores are really high, one or two really low)
skews - when's seomthings gone wrong , really extreme outliers
in exam, will be given a distrubution diagram, ask what type of distribution is this displaying - 'its this type of skew and this is why I know that'
'distrubution of data' this is what I should talk about
Correlational data - correlational coefficient
the closer the number is to +1 or -1 the stronger the correlation
symbols + or - telling whether correlation is positive or negative (positive - variables are increasing at the same rate and negative - one is going up and one is going down)
want number far away form 0 as possible
anything above 0.8 is a very strong relationship (both responding in exactly the same way) and 0.2 is a very weak relationship (something else causing the results to do as they are)
weaker, stronger positive/negative correlations
Content analysis and thematic analysis
content- analysis - converting the data into numbers
converting qualitative data to quantitive
comes up most
how to carry it out
define cateogries of analysis
create a coding frama (words, phrases, concepts, responses that could be found in the texts
developing rules for the coding - for each coding, clear operationalised list (stranger anxiety - crying, avoiding contact with the stranger etc)
Tally up each time one of your codes occurs in the text
count totals for each category, display these totals on a bar graph and draw conclusions
advantage - replicable, anyone can take the same coding frame and analyise
disadvantage - reductionist, simplifying data into words or phrases also subjective - researcher is picking what they're looking for, only what they want to find
thematic analysis - creating themes around the data to draw conclusions
keeping it qualitative, analysing general themes
comes up rarely
how to carry out
read the data, understand it, make notes
generate initial codes - create shorthand (initials) go through data and pinpoint when the codes occur
search for themes - use the data
review themes - pick out quotes that will support the theme, make conclusions - what does each theme capture about the data
conclude themes - conclusion what the themes suggest about the researcher question