Statistical Parametric Mapping (SPM) is a statistical method widely used in the analysis of brain imaging data, particularly in functional Magnetic Resonance Imaging (fMRI) studies. Determining the appropriate sample size for SPM analysis is crucial for ensuring that your study has enough power to detect a statistically significant effect if one exists.

Here are some steps and considerations for determining sample size for SPM analysis:

Regarding the use of sample size calculations from statistical analyses designed for discrete data:

It is not entirely appropriate to use sample size calculations from discrete data analyses directly for SPM analysis without adjustments. SPM and other neuroimaging data have unique characteristics, such as spatial correlation between adjacent voxels (autocorrelation), which must be accounted for in the sample size calculation. Discrete data analyses, such as those used in epidemiological or clinical trials, do not typically consider these spatial characteristics.

However, some of the underlying principles of power analysis can be applied. For example, the concept of effect size, significance level, and power are relevant to both discrete and continuous data. The difference lies in how these factors are estimated and applied in the context of neuroimaging data.

In summary, use sample size calculations that are tailored to the specifics of SPM analysis, taking into account the spatial and temporal autocorrelation inherent in neuroimaging data. If you are not familiar with these methods, it is highly recommended to consult with a statistician or an expert in neuroimaging analysis to ensure that your sample size is adequately powered for your study.