Data Recording, Analysis And Presentation Flashcards
How is raw data designed?
By using a tally short which summarises the data
What’s the use of raw data tables?
To get an overview of the data and see if there are any anomalies so the correct measure of central tendency can be used
Nominal level data
Categories of behaviour
Doesn’t tell us anything about the behaviour
Ordinal level data
Ranked order
Shows what ppts did the behaviour most or least
Interval level data
Level of measurement
Very Detailed as tells us most bout each ppt
Ratio level data
Type of interval
Data can’t go below 0
Primary data
Carry out own research and data is gathered directly from participants
Researcher knows exactly what controls were put in place etc but sometimes psychologists can’t carry out research themselves
Secondary data
Data that already exists
Good as can gather data that is too expensive to carry out by psychologists
May not be valid or reliable and there may be EV’s
Qualitative data
Data in form of words
Tells us experiences of ppts
Hard for comparisons to be made
Quantitative data
Data in the form of numbers
Allows comparisons to be made
Easily summarised
Reliability is easy to be found as results can be repeated
Measures of central tendency
Mean
Median
Mode
What measure of central tendency do you use with interval level data?
Mean
What measure of central tendency do you use with ordinal level data?
Median
What measure of central tendency do you use with nominal level data?
Mode
Measures of dispersion
Range
Variance
Standard deviation
Range
Subtract lowest from highest and add 1 or 0.1
What measure of dispersion do you use with interval level data?
Standard deviation and range
Variance
Measures how much a set of numbers is spread out
0 variance shows they are identical
A small variance shows they are close to each other
A high variance shows the data is spread out
Standard deviation
More detailed as it uses every piece of data
Ratios
A way of expressing proportions of a whole population
Frequency table
Shows the tally of behaviours
Pie chard
Show each behaviour as a proportion of the total
Line graph
Useful to show behaviour over time
Bar chart
Useful and meaningful way of presenting data
Histogram
Can only be used if the data is continuous (0-3 4-6)
Scatter diagram
Graph that represents a correlation
Psychologists can work out direction and strength of the correlation
They show any outliers
Proportionality
As one variable increases by a certain amount, the other increases by the same proportion
Inverse proportion
If one variables doubles, the other will half
Convert a fraction to a percentage
Divide top by bottom times 100
Convert percentage to fraction
Write down the percentage over 100. Simplify the fraction, if it isn’t a whole number, multiply top and bottom by 10 for every no after decimal point
To find the amount a percentage represents
Divide the amount/total by 100 and times by the percentage
To find the number represented by a fraction
Divide amount/total by bottom and times it by the top
Normal distribution curve
Data about behaviour from a representative sample of a target population will be a curve of normal distribution
A normal curve is when all the measures of central tendency occur at the highest point of the curve
Skewed distribution curve
When the distribution curve represents data from an unusual set of data (mean diff from mode)
Negatively skewed curve
Fewer people at the lower end and more at the higher end of the scores
Mean and mode are different
Positively skewed curve
Most the population have more low scores than high
Probability
Psychologists want to know what the probability is that the IV affected the DV, if there is a low probability they won’t believe it had an affect
Level of significance
The level that psychologists reject the null hypothesis is 95% (95% chance IV affected the DV and 5% chance it being due to chance)
If there was a need for the results to be as certain as possible (impacting society) it can be set at 99% or if the sample is small it can be set at 90%
What does p
Probability of results being due to chance is less than 5%
How do psychologists find out the significance level?
Using statistical tests
Inferential statistical tests
Chi square Binomial sign Mann Whitney U Wilcoxon signed rank Chi square Spearmans rho
Criteria for using a non parametric inferential test
What level of data is being used
Is it a test of difference or correlation
If it is a test of difference does it use am independent or repeated measures design
Criteria for using a parametric test
The data has to be interval or ratio
The data has to have a curve of normal distribution
The variances should be similar
Should be no extreme scores
Type 1 error
Accepting the alternative hypothesis when there is no difference
We believe IV has affected DV when it hasn’t
Behaviour is due to chance
Conducting to lenient of a test
Can avoid this by setting significance level higher (0.01) but this might cause a type 2 error
Type 2 error
Rejecting the alternative hypothesis and accepting a null hypothesis when the IV is actually affecting the DV
Being too strict and setting the significance level too high
What is raw data?
The data that psychologists collect for each participant