You are on page 1of 7

RM Cheat Sheet!

There is lots of stuff to remember in RM and sometimes having all of those little things in one
place is really helpful! And rather than me just write out what I think might be useful for you
guys, I thought we could try something a bit different!

Let’s make a collaborative ‘cheat sheet’ with all of the little things that it might be useful to
remember for RM - things like Excel shortcuts/formulae, SPSS hints and tips, all our silly
ways to remember bits of maths etc.

I have added a few things here already to get us started and you can suggest additions to
this cheat by getting in touch via email or on the discussion forum.

Looking forward to seeing what we come up with together! Hopefully this will be a resource
that will come in handy throughout the course!

Alex B
Useful Excel Formulae
Here are some formulae that you are likely to use regularly when working in Excel. Use the
link above to suggest additions!

Formula Description Example(s)

=average() Calculates the mean of the =average(5,10,15,8,4,5,6)


numbers/cells within the
brackets. =average(A2:A20)

=stdev() Calculates the standard =stdev(5,10,15,8,4,5,6)


deviation of the
numbers/cells within the =stdev(A2:A20)
brackets

=count() Counts the number of cells =count(A2:A20)


within a range containing
numbers

=countA() Counts the number of cells =countA(A2:A20)


within a range containing
anything (including text).

=sum() Calculates the sum of a set =sum(5,10,15,20)


of numbers or cells. (Adds =sum(A2:A20)
them up).

=sqrt() Calculates the square root =sqrt(4)


of a number or cell =sqrt(A5)
Mathematical notation
When typing formulae or calculations you can use the following symbols.

‘*’ The asterisk represents times/multiply. So to calculate 2 x 2 you would type 2*2

‘/’ The forward slash represents divide. So 4 divided by 2 would be written as 4/2

‘^’ This symbol means ‘to the power of’. You can find it with the 6 key on your keyboard.
You can use this symbol to square something e.g. 8^2 means 8 squared. 8^3 would be 8
cubed. 8^100 would be 8 to the power of 100.

‘<’ This symbol means less than. The number on the left is less than the number on the
right. E.g. y < 2 means y is less than 2.
Helpful tip: This symbol looks like an L! L = less than! <3 = heart emoji! ‘Less than 3’.

‘>’ This symbol means greater than. The number on the left is greater than the number on
the right. E.g. y > 2 means y is greater than 2.
Helpful tip: These symbols look like little crocodiles. The crocodile always wants to eat the
bigger number. So the ‘mouth’ of the symbol always points towards the bigger number.
Stats!
Errors in Statistics

Type 1 error - False positive


Type 2 error - False negative

Alpha and p-values

● Alpha in Psychology is 0.05 or 5%


● A p value is the probability that your result is simply due to chance.
○ p < alpha = significant (p is low the null must go!)
○ p > alpha = non-significant
● Smaller = significant
● Not-Smaller = Not significant

● Always remember that p should be reported to 3dp with no leading zero.


● When referring to significance we never say a result is ‘insignificant’ instead you
should say ‘not significant’ or ‘non-significant’.
Reporting Results Guide
When reporting results for practical reports or posters, you should generally follow
this ‘recipe’
Data Handling
● Say what the numbers are! I.e. How did you get from raw scores to the
numbers you use in your analysis.This is where you would tell the reader if
you have taken an average, or totals for example. You should also say here if
you have removed any outliers or done anything else to your data.
● E.g. The number of correct responses was added up to give a total accuracy
score per participant.
● Or E.g. The mean of the three trials were calculated to give an average score
per participant.
Descriptive Statistics
● This is where you should report your descriptive statistics.
● This should include a measure of central tendency AND a measure of
variance
● This is usually mean and standard deviation (or standard error or confidence
intervals).
● For some tests (particularly non-parametric tests) you may need to report the
median instead of the mean. With median you should report the Interquartile
Range.
● You can use a table OR a graph OR report the numbers in the text. Don’t
replicate data in multiple formats.
● If you use a table or graph don’t forget to format it using APA guidelines!
Which test and why?
● You should identify which test will be used to analyse your data here.
● This is where you can mention the assumptions tests – if you violate the
assumptions of a test you should report the outcomes of Shapiro Wilk or
Levene’s for example and say which test you will use instead.
● If your data meets the assumptions for the first choice of test (parametric test)
then it is fine to just say ‘the data met the assumptions of a …….. test’
Report the stats!
● For most stats tests you need to report the test statistic and the p values
● Remember that your p value should be reported to 3dp with no leading zero
and p should be italicized!
● See below for lists of the numbers you need for each test!
General Example for reporting stats
The score for IV Level 1 (mean = , SD = ) was higher than for IV Level 2 (mean = ,
SD = ). A name of statistical test here test revealed that this difference was
significant/non-significant (Test statistic etc, p).
What numbers do I need for each test?
Here you will find the numbers that you need for each test. I have grouped them in
parametric and non-parametric pairs where possible
___________________________________________________________________
Binomial Test
Descriptive Statistics: Proportions (either absolute or in percentages).
Test proportion (e.g. 0.50). This is the expected frequency you compare against.
p value for the binomial.
___________________________________________________________________
Chi-Square Test of Independence
Descriptive Statistics: Proportions (either absolute or in percentages). Use a
contingency table if you prefer!
(χ2(df)=......, p=.....)

Fisher’s Exact Test


Descriptive Statistics: Proportions as above.
p value only
___________________________________________________________________
Chi-Square Goodness of Fit Test
Descriptive Statistics: Proportions (either absolute or in percentages).
(χ2(df)=......, p=.....)
___________________________________________________________________
One sample t-tests:
Descriptive Statistics: Mean and standard deviation (or SE or 95% CIs).
Test value – this is the number you compared against.
(t(df) = ….., p = …… )

One sample Wilcoxon:


Descriptive Statistics: Median and IQR
(Z=…., p=……)
Here you should report the Standardized Test statistic from the Wilcoxon Test
Summary Table – this is the Z value. The p value is the ‘Asymptotic Sig’ value.
___________________________________________________________________
Independent Samples T-Test
Descriptive Statistics: Mean and standard deviation (or SE or 95% CIs).
(t(df)=….., p=…..)

Mann Whitney U-Test


Descriptive Statistics: Median and IQR
(U=…., z=..... p=….)
Note: Some online resources suggest you should report z with the U value, other
sources don’t. It is a good idea to make a note of it, but you won’t be penalized if you
don’t include this in a practical report for example. U and p are the most important.
___________________________________________________________________
Paired Samples T-Test
Descriptive Statistics: Mean and standard deviation (or SE or 95% CIs).
(t(df)=….., p=…..)

Wilcoxon Signed Ranks Test for Related Samples


Descriptive Statistics: Median and IQR
(Z=...., p=....)
___________________________________________________________________

Link to Which Test? Flow chart here

You might also like