Services
Statistical support from first question to final defence
Most analysis problems begin long before the analysis. We work through your study in the order a statistician would, so each decision rests on the one before it and you can explain all of them. Come to us at any stage; we start where you are.
Our approach
How we solve a statistical problem
Real research rarely fits a textbook recipe. We follow the stages of a statistical investigation set out by Chatfield (1988). Taking them in order keeps the analysis honest and your conclusions defensible.
In practice it’s a loop, not a line. What we see in the data can send us back a step, and we’ll tell you when it does.
After Chatfield (1988), Problem Solving: A Statistician’s Guide, Chapman and Hall
01
Understand the problem
Before we choose a method, we ask about your field, your objectives and how the results will be used. Then we restate your research question in statistical terms. The right answer to the wrong question helps no one, so this step comes first.
02
Check how the data was collected
How data is collected decides which analyses are possible. If you haven’t collected it yet, we settle sampling, sample size and your questionnaire first. If you already have it, we find out exactly how it was gathered before trusting any result.
03
Look at the data first
We check its structure and quality before any formal test: errors, missing values, outliers, and how each variable was measured. A careful first look often shows which analysis fits, and which ones don’t.
04
Choose the simplest analysis that answers the question
We match the method to your question and your data, not to what’s in fashion. We check every assumption. When the choice is close, we try more than one approach and see whether the conclusions still hold.
05
Interpret and explain
Output isn’t a finding. We compare your results with the literature, turn them into plain-language findings, and prepare you for the questions your supervisor, examiner or reviewers will ask.
What we do
Four ways we work with you
Approach steps 01–02
Research design
Sampling, power and instrument design, settled before you collect data so the analysis stands on something sound. This includes turning your topic into research questions and hypotheses you can test.
Research questions
Sampling
Power analysis
Questionnaire design
You receive
Research questions and hypotheses you can test
A justified sample size, with the calculation shown
An analysis plan matched to each question
Approach step 03
Psychometrics and measurement
Scale development, adaptation and validation. We test reliability and validity properly, then report them the way examiners and reviewers expect.
EFA
CFA
CR · AVE · HTMT
You receive
Item analysis with clear keep, revise or drop decisions
Reliability and validity evidence in a single table
Guidance on reporting it in your methods chapter
Approach steps 03–05
Quantitative analysis
From data cleaning and descriptive statistics to structural models. We choose the method your question needs, check its assumptions and explain why it fits.
Regression
Mediation
Moderation
SEM
PLS-SEM
You receive
A clean, labelled dataset with every change recorded
Annotated output and tables formatted for your thesis or journal
A plain-language walkthrough of every result, and practice with likely examiner or reviewer questions
Alongside any stage
Workshops and training
Hands-on sessions for supervisors, postgraduate students and research teams. Participants work with real data and leave able to run and explain the method themselves.
Workshops
PLS-SEM
Formats
Half-day introductions
Multi-day intensives
Ongoing clinics for research groups
Get in touch
Have a dataset and a question? Let’s think it through
Tell us about your study and your deadline. We’ll reply with a scoped plan and a clear quote.