Can an economic structural model support hypothetical and experimental evidence ? Preference parameters before and after the Great East Japan Earthquake 1