Data Analysis
Perform Lit Search for how data is normally represented
Is the Raw Data Normally Distributed?
No
Yes
Is this how the data should be represented?
Yes
No
Use non parametric plot
FIRST OPTION: Reorganize the Data
Find the Source of Variation (Scatterplots and Histograms)
If data is skewed
If Outliers are found
Can/Should outliers be removed?
Yes, there are enough subjects in the study for the power analysis
No
Were outliers different in any way NOT related to the study?
Normalize data to each subject
Test again for outliers
Explain cause of variation in paper. Remove outlier. Restart
Is there reasonable way to group subjects to explain variation?
e.g. Weight
e.g. Sex
Find a relevant ratio for a linear transformation
Test again for normality
Go to statistical analysis
SECOND OPTION: Transform the Data
Ways to Look at and Interpret Data
Group Subjects by:
Sex
Treatment
Timepoint
In each group, do subjects show a change over time?
Do males and females perform differently either befofe and/or after treatment?
Perform any common normalizations or transformation according to lit search
For proportions or Percentages
When looking for a change from normalcy, not a specific increase or decrease
For skewed data
Negative Skew
Positive Skew with non-negative values
Logarithmic transformation (for cumulative, multiplicative effects or exponential data.
Exponential transformation (for log trends e.g. decay)
Power transformation (for power components e.g. area, volume)
Arcsin transformation
Remove and restart
Square Root transformation (for counts or data with a power, e.g. area, volume)
z-score transformation
How does recovery vary across time based on treatment? Were there any differences between groups before treatment?
Statistical Analysis
Group difference
Time/Mixed effect
Distribution difference
Prediction
Sample <10 and/or Normality fail
YES - Non parametric test
NO - Parametric and Non-parametric
Chi-square test - normality
Levene's test - variance
ranksum - two sample
signrank - paired test
Kruskal-Wallis - N-sample
t-test - two-sample
paired t-test - paired test
ANOVA - N-sample
Kolmogorov-smirnov test - normality
rmANOVA - Nsample/repeated
Chi-square
Exact Fisher Test
Linear Classifier (LDA)
Support-Vector Machine
k-Nearest-Neighbour
Artificial Neural Network
Data Reduction (Dimension reduction)
PCA
LDA
ICA
multicolinearity
Backward feature elimination
Decision tree/random forest
Multiway array decomposition
Plot Data for exploration
Non parametric plot
Box-whisker plot - group spread/outliers
Scatter plot - data spread/individual
Histogram - Distribution
Parametric plot
Mean + S.E.M. - group difference/spread
click to edit
Use parametric plot + scatter plots
Explore data transformation