13 - Statistics – organising the data Flashcards
Raw data/ scores
Untreated, unconverted values obtained directly from measuring process used in a study.
Categorical variable
Variable where cases are merely placed into independent, separate categories.
Star sign
JUST ONE category.
Measured variable
Variables where cases measured on it are placed on some sort of scale that has direction.
Any other sort of variable above the level of mere categorising.
In many experimental studies
Independent variable = Categorical Variable
Normal or hot room = categorical
Dependent variable = Measured variable
Aggression rating = measured
Levels of measurement
Levels at which data are categorised or measured.
Nominal measurement
Level of measurement at which numbers are only labels for categories
Associated with categorical classification system.
“Tall” and “Short”
Ordinal measurement
Levels of measurement at which cases are arranged in rank positions. (ranking scale)
Tallest to shortest
Gives more information than “Nominal”, but still lacks a lot.
Interval
Level of measurement at which each unit on a scale represents an equal change in the variable measured.
Measures every person.
Coding (nominal)
Giving “dummy” numbers to discrete levels of an independent variable. (because of system)
Example: studying education level.
High School = 1
Bachelor’s degree = 2
Master’s degree = 3
Frequent data/ frequencies
Numbers of cases in specific categories.
Example: 10 people said High School = 10 Frequency data ^
Quasi-interval scale (interval level of measurement)
Scale that appears to be interval but where equal intervals do not necessarily measure equal amounts of the construct.
Example: IQ scale. 120 IQ vs. 100 IQ, does not mean the other person is 20% more intelligent. ?
Not calibrated from a Zero-Point
Ratio Scales
interval-type scale where proportions on the scale are meaningful; usually an absolute zero exists.
Example: weight (kg)
Calibrated from a zero-point
Changing data from one level to another
You can only go downwards.
Median split method
dividing a set of measured values into two groups by dividing them into high and low at their median.
Example: Having done an interval score on people anxiety and finding the median (average) and placing everyone above that in “high” category and everyone below in “low” category.
Continuous scale/ variable
scales where there are no discrete steps; theoretically, all points along the scale are meaningful.
Example: age
Discrete scale/ variable
Scale on which not all subdivisions are meaningful; often one where the underlying construct to be measured can only come in whole units
Example: number of children.
Data set
Group of data points or values which can be summarised or analysed.
Example: Premier League Table
Central Tendency
Formal term for any measure of the typical or middle value in a group
Mean
Median
Mode
Dispersion
technical and general term for any measure of the spread of values in a sample of data or population.
Basically: not only showing the average or median, but how much the numbers vary. All the grades amongst the different grades.
Mean (arithmetic)
Average of values found by adding them all and dividing by the number of values in the set.
How to calculate grade average
Outliers
“Rogue” values at an extreme distance from the centre of the data set. Can be removed from analysis, but must be reported.
Trimmed Mean
The mean of a data set with its most extreme 5% of values removed.
Example: Removing 2.5% of the highest and 2.5% of the lowest score.
Median
Measure of central tendency; middle value of data set.
Basically: sort numbers (rising) to find the middle. If two in middle. Add and divide by 2
Example: 2,4,7,9,12
Median position/location
Position where median is to be found in an ordered data set.
Basically: just the position. Number 5 in a row of 10
Class Intervals
Categories into which a continuous data scale can be divided in order to summarise frequencies.
Example: 0-10 years, 11-20 years etc…
Model/modal value
Measure of central tendency – most frequent value in data set
Example: if 5 is the most said score on a movie, the modal value is 5.