Show Menu
Cheatography

PSY2206 Methods and Statistics Cheat Sheet (DRAFT) by

Key SPSS functions and terms to support PSY2206 exam

This is a draft cheat sheet. It is a work in progress and is not finished yet.

Data terms and Inputs

 

Glossary terms

Descri­ptive Statistics (DS)
describe data set concisely and in a repeatable format
Central tendency (DS)
several different types- all attempt to give a single number that represents your variables. most common is average but also includes mean, median and mode
Dispersion (DS)
aims to give a single number that represents the spread or variab­ility of the variables. The larger the dispersion value = the larger its variab­ility. standard deviation is the most common measure
Standard deviation (DS)
an estimate of the average variab­ility of spread of the variables.
Standa­rdised scores (DS)
is measured in terms of standard deviations and allows instant compar­isons to the means score on a variable- can then tell if score is greater or lesser than the mean.
Infere­ntial Statistics (IS)
allows you to make inferences from your sample about the larger popula­tion. Will tell you the probab­ility of your results occurring in the popula­tion. However it does not tell you anything regarding the size of the differ­ence.
Effect sizes
indicate the strength of the relati­onship between your variables #. should be reported in conjun­ction with the probab­ility level associated with any infere­ntial statis­tics.
Parametric variables (IS)
tests that assume that the data approx­imates a certain distri­bution e.g. normal distri­bution. Allows us to make inferences that are capable of producing accurate results. Must double check assump­tions.
Non-pa­ram­etric variables (IS)
tests that make fewer assump­tions about the data and thus can be used to analyse a more diverse range of data. Less powerful than parame­tric.
Indepe­ndent Groups Design
looking for differ­ences on a variable between separate groups of people. two groups often use t-test or non-pa­ram­etric Mann-W­hitney test. When invest­igating differ­ences between more than two groups the parametric betwee­n-g­roups ANOVA- often follow up with a post-hoc test
Repeated Measure Design
looking for differ­ences on a variable within the same group of people- look to see have scores have changed. If you have tested your partic­ipants twice you can use a paired t-test or non-pa­ram­etric Wilcoxon test to identify if the difference is signif­icant. If you have tested them more than twice use a parametric within groups ANOVA.
 

Identi­fying types of data

Variable
what we measure- must identify: Type (discrete or contin­uous), Level of measur­ement (normal, oridnal, interval or ratio), Role in the research study (indep­endent, dependent or covari­ate).
Data
inform­ation we collect about the variable
Data Set
collection of inform­ation about the several variables
Discrete variables (Categ­orical varaia­bles)
variables that contain separate and distinct catego­ries. e.g. Gender- split into female, male etc.
Continuous variables
not spilt into distinct categories e.g. age. fractions are meaningful
Measur­ement properties
magnitude (can order the values in a variable from highest to lowest), equal intervals (unit difference on a measur­ement scale is the same regardless of where that unit difference occurs on the scale e.g. temper­ature) , true absolute zero (zero point on the scale is the point where nothing of the variable exists and there are no scores).
Level of measur­ement
Nominal, Ordinal, Interval and Ratio
Nominal
Variable has none of the measur­ement proper­ties. use the numbers in the variable only as labels
Ordinal
have the measur­ement property of magnitude- ordere­d/r­anked
Interval
measurment properties of magnitude and equal intervals.
Ratio
measur­ement properties of magnitude, equal intervals and absolute zero.
Covariates
broad label for a variable in a research design that's neither an IV or DV. Can be used to take account of other factors that might influence the relati­onship between IV and DV. Can also exist in research designs where no IV or DV exist.
 

Inputting Data- Variable View Window

1) Identify your variables and label
choose concise and distinct variable names e.g. Time, Age ect.
2) Decide on variable type
Variable view window (bottom left) > Numeric (can change to things such as dates ect.) in the values section you can click the three dots and assign numbers to scores e.g. 0=easy
3) Assign level of measur­ement
variable view window > measure > scale/­nom­ina­l/o­rdinal

Inputting Data- Data View window

Sorting cases
Can be a useful tool in ordering variables you have inputted (e.g. highest to lowest). Data > Sort cases > select variable you wish to sort > select either ascending or descending > Ok >
Recoding variables
converting the data in a variable into new data (as a new variable with different meaning or measured on a different scale). e.g. wanting to examined differ­ences in exercise taken less than 12 months after transplant and more than 12 months after- can recode the data so it establ­ishes these different groups. Transform > Recode into Different Variab­les... > select the variable (e.g. time) and move to right box > Output variable box allows you to name new variable > press change > Old and new values.. > input the values from the existing variable that you want to recode in the left hand box > new value that you want to apply in the value box under new value > add > click continue after adding all the values > can then go to variable view window and add labels

Measure of Central Tendency

Central Tendency (CT)
a repres­ent­ative value from the data- a summary of all the values on a variable in your data set. Always refers to a single variable- thus if you have several variables in the data set they each need their own CT
Dispersion
how much the data deviates from the central tendency
Mode
simplest measure of central tendency to calculate- most frequently occurring. Order data if doing it manually. SPSS: Analyse > Descri­ptive Statistics > Freque­ncies > move variable containing your scores into the variables window >