This podcast answers the questions raised below, by a client trying to understand how to develop hypotheses in a fairly difficult case. The case looks at ways to increase the US Postal Services market share and profitability. The client has graciously allowed this one detailed response and his notes to be shared in our public section.
“A big overarching question I have with the answer first approach is I don’t know if my initial hypothesis should be broad (in which case they cannot be tested with data unless the hypotheses are first refined and narrowed, so I could only ask what areas I would explore to do that) or alternatively to make my initial hypotheses narrow and precise (in which case I can ask for data to test them but if my hypothesis are wrong then I will be stuck). Any thoughts?”