This is analysis of some biosci/biomed screening data. 2 subsets to be analyzed, each with <75 test subjects and each having a singular outcome. initial analysis needed for each is a linear regression using logdose and age at testing. then a regression that includes interaction analysis for age at dose (ages split into first a 2-way, then a 3-way analysis each with categories pre-determined). We'd then like everything repeated with dose in place of logdose to evaluate difference.