Ipsative instruments are generally seen as violating the assumptions usually required for parametric linear regression (OLS).  Because the C-VAT instruments use a forced choice format for theoretical and practical reasons (our ipsative instruments are highly predictive), we use a proprietary cluster analysis with nonparametric measures of association for most of our research (see Nelson and Gopalan in Organization Studies, 2003 for more information).  Cluster analyses are more forgiving of most distributional and other assumptions than OLS techniques, as are gamma and lambda coefficients which are used in a manner analogous to traditional correlation coefficients.  Both lambda and gamma are proportional reduction of error measures and are analogous to R squared in their interpretation.  

Below are contingency tables produced by summing counts of people belonging to mechanically generated clusters using the 80 items generating the 16 scores on individual PVP profiles.  The first data set is from the IT division of a large US utility:

Cluster    I            II
Hierarchical Level
1                                                     6             4
2                                                     4             5
3                                                     5            12
4 (Staff)                                          1            21

Chisquared:                              P

As can be seen, membership in cluster one drastically increases the likelihood of occupying a managerial position in this division.  Gamma = .77, Lambda = .77 or .54 depending of the conservativeness of the calculation.  This is a very strong result, far in excess of traditional psychometric instruments.  As baselines, consider that the association between membership in the Republican Party and a vote for George Bush in the last US presidential election was lambda= .66, gamma .54.  As another example, the lambda for the association between education and income in the US is  .60, the gamma is .53.

Another example from the US taken from the US’s largest charity and NGO is similar:

Cluster    I               II                III
Hierarchical Level 
1                                                  5                3                1
2                                                   4                2                7
3                                                  5                 7               6
4  (Staff)                                      1                 8              13

A conservative calculation of lambda is .64, gamma is ,58.

Changing the cultural setting, the same analysis was performed at the headquarters of a large Brazilian retailer with the following result:

Cluster                                          I         II
Managerial                                  10         4
Nonmanagerial                              8        21

Lambda=  .59
Gamma=   .57

Similar association are obtained for hierarchy  and involuntary turnover rates in upper management of a large Brazilian appliance manufacturer, incarceration for white collar crime in the US, sales productivity in Brazil, and for hierarchical level of executive MBA students in Hong and Singapore.

Analyses of corporate culture using the CVP instruments yields similarly high associations on a range of independent variables from satisfaction levels, gender, function, seniority, social network position, conflict levels, turnover, industry, and performance, including bankruptcy.  C-VAT can assist clients in determining the predictive validity of their HR Objectives.