Questionnaires and organizational hierarchies

Survey questionnaires and graphs

“…His optimal questionnaire has a minimum number of expected questions, a measure of the efficiency with which the information is collected, defined in Picard (1965). Picard's algorithm for obtaining an optimal questionnaire is similar to the algorithm devised by Huffman (1952) for use in coding theory, which is also discussed in Parkhomenko (2010).…”

“…However, he suggests no systematic way in which flow charts could be used in designing questionnaires. Parkhomenko (2010) modeled organizational hierarchies (and questionnaires) as graphs, with managers viewed as nodes of these graphs. Picard's technical restrictions on moving nodes with their branches within the graph are not required here.…”

“…In Elliott (2012), graphs are used mostly for testing electronic questionnaires. Other than the work of Parkhomenko (2010), little seems to have been done on the important topic of systematically designing a questionnaire which incurs a minimum response burden.…”

“…In this paper we introduce survey charts, graphical tools originating from survey practice, which we use to represent and transform survey questionnaires. Although our set-up is close to that of Picard (1965) or Parkhomenko (2010), our approach differs from theirs in some fundamental ways, specifically in the purpose of our research, the definitions that we give, and the procedures that we propose and follow. We succinctly present these differences here, with details given in section 7.…”

“…We succinctly present these differences here, with details given in section 7. Our survey charts are graphs with special properties, and cannot generally be identified with trees (the type of graphs used by Picard (1965) or Parkhomenko (2010)), either before or after we apply our transformations. We endow each survey chart with a finite probability space, richer than Picard's (1965) in that it has "points", or analytical outcomes, which capture the entire information collected by the questionnaire and are indispensable for a complete analysis.…”