An Experimental Assessment of a Newly Revised Democratic Maturity Test for High School Usage - Brief Article - Statistical Data Included

Journal of Instructional Psychology, June, 2000 by Robert C. Reiger

This is an experimental assessment of the present status of The Democratic Maturity Test (DEMO) for use with high school students. It is clear that the mean for part score entitled "Self-esteem" is too high, and must be revised immediately. All of the other scores seem to be functioning properly.

The Democratic Maturity Test (DEMO) is being developed for use in the assessment of individual's readiness for successful living in a democracy. It is based on the theory described by John Dewey that "a democracy is the interdependence of independent individuals." It has 200 true/false type items distributed in eight part scores, with the first four dedicated to assessing independence as "Personal Maturity," and the second four for measuring interdependence as "social Integration." The DEMO test has gone through more than a half dozen different revisions, and 7 of the 8 part scores appear to be in the final stage of such assessment.

High School Population

The high school assessment population included 121 senior high school students from a typical high school in New Jersey ranging In age from 15 to 19 years with a mean age of 15.95. and a standard deviation of 1.06 years. It included 66 female and 55 male students. There were 70 9th grade students (freshmen), 30 10th grade students (sophomores), 15 11th grade students (juniors) and only six 12th grade students (seniors) involved.

Validity of Student Scores

The Discerning Score is comprised of 21 pairs of test items that are either opposite to each other, or not compatible. A score of 15 or higher is considered to represent invalid test data. Not a single student obtained an invalid score, and characteristic erasing of scores in student records suggest serious test taking.

Independence of Eight Part Scores

First, and foremost we are concerned about the independence of organization of the eight part scores, and the degree to which they might overlap in terms of their functional content. A principal component factor analysis, using a constant was computed using varimax rotations with rotated factors as depicted in Table 1, below. There is a total independence of the eight part scores, each with a loading at the 900 level and with not a single significant contamination evident--scores of 300 level or better. The total variance is equally distributed among the eight part scores: 12 in each one.

Table 1

Principal Components Factor Analysis of DEMO Part Scores (N = 121)

                                                   Factors
Scores                       I        II        III       IV

Conformity         CFM      948(*)   -122       062       110
Assertiveness      ASS     -116       964(*)    058      -020
Sympathy           SYM      058       057       967(*)    061
Efficacy           EFF      103       019       061       970(*)
Caring             CAR      107       138       134       084
Locus of Control   LOC      168       098      -001       111
Coping Skills      COP      146       195       169       070
Self-esteem        EST      122       014       149       177

Scores              V        VI       VII       VIII      % TV

Conformity         104      170       136       114       12.52
Assertiveness      126      093       172       013       12.67
Sympathy           122      001       148       129       12.69
Efficacy           077      105       061       152       12.69
Caring             937(*)   079       154       191       12.47
Locus of Control   077      950(*)    119       163       12.58
Coping Skills      160      129       924(*)    107       12.21
Self-esteem        201      179       107       921(*)    12.19

(*) = Identifying Loadings.

%TV = Percentage of Total Variance.

Distribution of Means for Part Scores

It Is clear from the means for each of the eight part scores as depicted In Table 2 below that the mean for Part I--Self-esteem is far too high for high school students to be an effective instrument, and that it has to be revised. The mean of LOC is a little high but is still acceptable. The Standard Deviation is about average for LOC. Of all 8 part scores, the LOC score is the most technical, and deals with scientific decision making.

Table 2

Means for 8 Part Scores (N = 121)

        EST     COP     ASS     LOC     CFM     SYM     EFF     CAR
Mean   80.96   43.40   57.12   70.07   57.26   63.40   57.32   60.89
SD     14.07   11.54   11.46   12.72   12.02   13.80   11.81   14.32

Score Reliability Without Correction

The part score reliability for all 8 part scores without correction, based on correlation with total score are all very acceptable, with none of them less than 0.571 as shown in Table 3 below.

Table 3

Reliability of Part Scores Uncorrected (N=121)

        EST    COP      ASS     LOC     CFM    SYM     EFF     CAR
Corr.  0.696   0.719   0.571   0.712   0.621   0.653   0.600   0.716

General Findings

The general findings from the DEMO test at the high school level are as follows:

1. A t-statistic between means showed no significant gender difference in the mean scores for the male and female students.

2. An analysis of variance between the four different grades (9th through 12th) for the mean of the 8 part scores showed no significant difference between the grades for gains or loss in democratic maturity.