Vector Approach for Analyzing Survey Questions


  1. Villeneuve, S.
  2. Farrell, P.S.E.
Corporate Authors
Defence R&D Canada - Toronto, Toronto ONT (CAN)
A vector approach is presented here to analyze survey questions. The survey questions are designed as bi-polar constructs. Participants rated the usefulness, ease of use, and look and feel of six tools on a scale from 1 to 7. The vector method treats each answer as an element in a p-dimensional vector, where p corresponds to the number of participants. The response vectors are then compared to a reference vector. The percentages of usefulness, ease of use, and look and feel are calculated using the projection of the response vectors onto the reference vector. A clear advantage of the vector approach over statistical methods is that it does not invoke assumptions typically associated with those statistical methods. Thus the vector method can be applied to a wider variety of data sets, and the results are non-ambiguous compared to the statistical techniques. While the statistical method indicates that the participants seemed to have interpreted the three questions equally,the vector method shows clearly that the questions are not coincident and care must be taken in drawing the right conclusions. Other results specific to the experiment described herein are that the highest rating from the participants was for the Info Workspace tool and the lowest was the EBP tool suite. Also, Common Intent depended on variables other than technology. Nevertheless, the technology should be optimized to eliminate it as a potential detrimental variable.

Il y a un résumé en français ici.

Report Number
DRDC-TORONTO-TM-2005-151 — Technical Memorandum
Date of publication
01 Aug 2005
Number of Pages
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: