Isabel B. Myers and Katharine Briggs devised the Myers-Briggs Type Indicator® personality inventory in the early 1940s to assist people in developing self-understanding and enhancing their appreciation of others. They based their approach on theories developed by Swiss psychiatrist Karl Jung who believed that all humans have the capacity to observe and to organize information, but that there are natural, inborn differences in the ways people prefer to use these capabilities. Simply stated, the MBTI® expresses these basic mental tools as perception, decision-making, orientation and structure. Jung professed that each of these basic tools contained opposite processes of organization and structure, and that individuals would naturally select one or the other processes for each basic mental tool. In Jung's toolbox, Perception either used a process of Sensing (S) or Intuition (N); Decision-making used either Thinking (T) or Feeling (F); Orientation utilized either Extraversion (E) or Introversion (I); and Structure involved either Judgment (J) or Perception (P). Individual preference combinations of opposite factors within the four basic tools yield sixteen possible MBTI® types (Kirby, Linda K., 1997). Within these types Sensing (S), Thinking (T) and Judging (J) factors are the majority preferences in upper management, however no one type stands out at the top (Walck, Christa L., 1997). Thinking (T) and Judging (J) factors are found in the large majority of management across a variety of cultures and types of organizations (Kirby, 1997).
This research did not generate a formal MBTI® analysis for the participants, however a general distribution of traits based on the simple inventory questionnaire contained in Section 'A' was performed. This basic screening was designed to take advantage of the population surveyed to shed some light on the distribution of MTBI® factors among the surveyed population, and to confirm a general Thinking (T) and Judgment (J) distribution in excess of 50% to confirm to the researcher that the respondents viewed themselves as having characteristics typical of the majority of upper level management.
Section 'B' was the targeted test section. Data were gathered by allowing survey participants to choose a response to each statement on a Likert Scale ranging from Strongly Disagree (1) to Strongly Agree (5). Developed by Rensis Likert, Likert scales are useful tools for gaining insight into respondent's feelings, opinions and attitudes (Likert, 1932).
Although they offer the advantage of measuring a range of responses, Likert scales suffer from two basic problems. The first is the tendency of some respondents to assume a neutral or "fence-sitting" position on some, or all, issues. The second is the difficulty in determining the relative strength of conviction for each response, as the response values of 1, 2, 3, 4, 5 are ordinal rankings with no corresponding interval and ratio scales (Shiken, 2000). The first problem was effectively minimized by selecting survey groups composed of experienced senior managers. Senior public sector managers serving municipalities with populations larger than 75,000 have served an apprenticeship in several lower level management positions. Likewise, senior private sector managers at the vice president level and higher have also progressed up the ranks through various lower level assignments. This extensive career experience, coupled with guaranteed anonymity assured by the random selection process, the absence of personal information from the survey instruments and the pre-addressed metered return envelopes for completed surveys minimized the probability of a default neutral ranking within either group.
The second problem was addressed by accepting the ordinal nature of the responses, and applying statistical analysis focused on location of response. Due to the discreet nature of the observations, the data were summarized as counts occurring in the various categories, as suggested by Clason and Darmody (Clason and Dormody, 2001).
Group means were compared rather than individual responses to avoid the risk of assuming too much linear scale with a resulting loss of empirical meaning or scientific significance (Adams, Fagot and Robinson, 1965). Differential research methodology was applied to group data for Section 'A' and 'B'. Section 'A' was analyzed by a simple distribution chart and a calculation of the percentage of respondents identifying themselves as conforming to Thinking (T) and Judging (J). Section 'B' is the values section and the null hypothesis was tested for the mean response for each value statement. Section 'B' was analyzed using one-tailed t-tests for two samples with unequal variance to determine the degree of correlation between group means. The null hypothesis confirmed that there was no significant difference between the two population means. Significance was established at .05, with values below that rejecting the null hypothesis and values above that confirming the null hypothesis. If the null hypothesis was confirmed, it was assumed that both sample population sets utilize a similar value distribution in developing soft facts pertaining to the ranked statement, with observed differences in outcomes from analyzing similar inputs based on other criteria, such as their own personality traits and experience. If the null hypothesis was rejected it would indicate that, in addition to any other differences that might exist between individuals, significant differences exist in the underlying values and assumptions applied to the ranked statement as applied to the foundation upon which each sector builds its goals and objectives.
A primary component of the survey analysis was those statements that reflected each sector's view of the other sector's priorities. A positive correlation through the entire set of values measured confirming the null hypothesis would predict that there is good understanding between the sectors and that negatives arising from time to time stem primarily from the noise of competition within the public arena. Rejection of the null hypothesis would tend to indicate that there is an underlying value conflict between the sectors insofar as that particular statement that may create distractions beyond the noise of competition and make the movement to public-private partnerships inherently more difficult without the use of compensating mechanisms.
If the null hypothesis was rejected, interviews were performed with participants from each group again selected at random. These interviews focused on gathering information on the possible underlying causes of the measured differentiation, and options for effective compensating mechanism within the process.