Welcome!
This notes pack has been laid out to make it as easy and stress free to get through this stats
course
Some things to note:
Each different test type has one sheet with explanation and steps, and another with
1 practice examples
The psych department has also provided further examples on vula. Use them!
The steps for each test do not use real numbers, so some of the cells may show error
2 codes like this
Err:509 #VALUE!
If you click on the #NAME? cell above you can see the formula in the formula bar
3 Sometimes I have added extra information in notes
These notes often explain the formula used, and where you find each value for it
hover over the red triangle
4 When in the test, DO NOT USE THIS EXCEL DOC AS YOUR WORKBOOK
UCT will be able to see who the original author is, and everyone who bought these
notes will have exactly the same formatting
YOU WILL BE CAUGHT FOR PLAGIARISM IF YOU DO THIS
Instead:
Open up a new excel workbook to use for the test
Go step by step for each question, and just refer to this for formulas or for
the step by step order
DO NOT GIVE MUYA THE SATISFACTION!
I know to not use this notes pack as my workbook in the test
Signed Date
Good luck! If you have questions or notice that something is missing before the test,
please let me know. Stats can be confusing, but is quite easy when followed step by
5 step. I will not be contactable during the test
Test guidelines
all answers must be to three decimals (including scientific notation)
You must use a full stop (.) as the decimal point
,Some main concepts
standard error how much we expect a mean will vary across samples of popn
if estimate falls outside of margin of standard error > significant
smaller sample = more error
standard error = standard deviation of sample means
test statistic (for Chi, T, F) what you expect to see under conditions of randomness
test statistic = residual / standard error
test statistic = (sample mean - popn mean) / standard error
null hypothesis the sample is the same as the population at large
alternative hypothesis changes in data are caused by independent variable
effect size tells us whether observed effects are practially significant to us
Symbol Glossary
α Alpha, Type I error rate
β Beta, Type 2 error rate
μ Mu, the population mean
σ Sigma (lower case), the population standard deviation
Σ Sigma (upper case), the arithmetic summation operator
χ2 Chi-square statistic, or Chi-square distribution
σ^2 Sigma square, the population variance
D Difference between two scores
E Expected frequency
F F distribution, or F ratio
k Number of groups in a design
MS Mean Square
N Population size
n Sample size
O Observed frequency
p Probability
Q Tukey’s Q statistic (studentized range statistic)
s^2 Sample variance
S²x Standard error of the mean
SS Sums of squares
The t statistic, used to test hypotheses about mean differences; also the probability
t distribution
x Sample mean
Types of variables
Types of variables
discrete (certain values in range) continuous (any value in range)
Nominal Ordinal Interval Ratio
scales of measurement
Ranked data (have
order but intervals Differences are valid, Ratios are valid, has
Categories are meaningless) no absolute zero real zero point
temperature, IQ
hair colour, religion grades, likert scales scores age, length
,scales of measure
No maths operations plus and minus
apply Only < or > apply apply Can do +, -, x, /
, Which test is which?
t test
Continuous data
Chi square
categorical data