why is this interesting? where are your research questions coming from
Research questions
formulate your research question in clear natural language
give them recognizable names if there are several or number them (like H1, H2 etc.)
you will restate these research questions more clearly after you have specified the design (see below)
Design of the Experiment
General remarks about the design
what kind of a study is it (e.g., a 2x2 within-subjects factorial design)
Sampling plan
how many participants will you recruit, and how, where, when?
any special restriction as to who qualifies?
Materials
what materials will you be using? how where they produced
you could include a link to the material if it is available online
Procedure
describe the experiment here pretty much like you would in a research article without space constraints
similar to the statement in the Mental Rotation Task description, but possibly already in full prose
what is the structure of the experiment?
what types of trials are there (critical, filler, …)?
how is each (type of) trial structured? (e.g., first fixation cross for 50ms, then …)
Measured variables
based on the design you described, what are the variables that you are going to measure?
how will they be measured in the experiment?
how will they be treated (e.g., rating scale data will be treated as an ordinal variable; or: XYZ is a factor with 2 levels (A and B) where A is the reference level in dummy coding)
Analysis Plan
Exclusion criteria
what are the criteria based on which you would exclude data from the analysis
for single trial data
for data from a whole participant
…
Confirmatory hypothesis testing
if applicable, describe any transformations that you might want to apply to the data
describe your statistical model
e.g., we will use the ‘brms’ package to run a Bayesian regression model regressing XYZ against A, B and C and their interactions
describe by what means you will test the hypotheses mentioned earlier
when you do this you will want to reformulate the hypotheses in a more precise fashion
e.g., if H1 is true, we expect that parameter X is credibly bigger than zero in the posterior distribution