Research methods and statistics Flashcards
Nominal data
When the DV is the number of participants in each category
Ordinal data
When the data is a rank, position or rating
Interval data
When the data is a real number/ measurement
What is a measure of central tendency?
A measure of where the centre of the data is.
Name the measures of central tendency
Mean, median, mode
What are measures of dispersion?
Measures of how spread out the data is.
Name measures of dispersion
Range, standard deviation
Advantages and disadvantages of central tendency measures
Mean- most sensitive, but can be distorted by extreme values
Mode- Can easily identify most frequent value, but sometimes there is not a common value
Median-Not distorted by extreme values, but not good for small data sets
Advantages and disadvantages of measures of dispersion
Range- quick to calculate, but doesn’t take into account all of the data
SD- most sensitive, but time consuming to calculate
Define quantitative data
Data that is numerical or categorical, it has values
Define qualitative data
Data with no numerical or categorical value, focused on detail.
Normal distribution
Bell shaped curve, mode median and mean all in the same place
Skewed distribution
Distributions leaning to one side. Mean median and mode in different places
Positive skew- tail to the right- mean and median higher than mode
Negative skew- tail to the left- mean and median lower than mode
Type 1 error
Too lenient and reject null- false positive- due to low significance level (e.g 90% or 0.1)
Type 2 error
Too strict and accept null- false negative- due to high significance level (e.g 99% or 0.01)
Chi squared test requirements
Independent groups, difference, nominal
Sign test requirements
Repeated measures, nominal, difference
Spearman’s RHO
Correlation, ordinal (or both ordinal and interval)
Pearson’s R
Correlation, not ordinal
Mann-Whitney
Difference, ordinal, independent
Wilcoxon
Difference, ordinal, repeated measures
Unrelated t-test
Difference, interval, independent groups
Related t-test
Difference, interval, repeated measures
When to use tests with interval data? (Parametric)
When data is normally distributed and standard deviations are similar. Can use parametric tests usually anyway unless it is clear in q that data is skewed. Then use ordinal data tests.
Formula for accepting rejecting hypothesis
State critical value and why (e.g, is two tailed, 5% significance, N=11).
State if observed level is lower or higher.
Therefore results significant/ not.
Accept or reject.
Is matched pairs repeated or independent measures?
Repeated.
Chi squared table name?
Contingency table
Df for chi squared?
(Columns of raw data (NOT TOTALS)-1) x (Rows of raw data-1)
Df for Pearson’s R
Number of participants-2
Df for unrelated t-test
Df= N1+N2-2
Df for related t-test
Df= N-1