I'm just reading some papers on genome-wide association studies and find that most of them use SNPs as a marker. I understand that they're the most polymorphic and thus the best for accurate correlations, but would microsatellites provide another viable method to check for genetic correlation? They're highly polymorphic and small enough to showcase differences, while maybe allowing for more statistical significance.
Microsatellites could be used for GWAS. Actually they were the basis for linkage studies and were also used in the first association studies. The main reason why they have been replaced by SNPs is that in the human (or other organims') genome, there are by far more SNPs than microsatellites. GWAS rely on linkage disequilibrium between the marker and the real causal variant. LD decreases quite fast with distance, thus, the more markers you use, the more likely you are to genotype a marker in LD with the causal variant. In the end, I would strongly recommend to use SNPs to perform GWAS.