38
NewDirection Program Evaluation
NewDirection Program Evaluation
Recidivismhas been a big problem among juvenile offenders in the USA. In a bidto reduce the recidivism among the juvenile offenders, the CountyJuvenile Court and Department of Corrections over the past few yearshave implemented a “New Directions” program. This program wasdeveloped to serve juveniles between the age of thirteen years andseventeen years. The New Directions program is a oneday program,which involves only selected inmates who undergo screening andtraining. Despite its success, some people have questioned itseffectiveness. This necessitated a study to be conducted to evaluatethe effectiveness of the “New Directions” program in reducing therate of recidivism among the juvenile offenders. A further analysiswas also conducted to determine if the effectiveness of the programdiffered across races, the different ages, and the type of offencethe juvenile committed.
ResearchMethodology
Simplerandom sampling was used to select the juveniles who were toparticipate in the study. A sample of 300 juveniles who had appearedbefore court and placed on probation in 2012 was used, with onehundred and fifty of them participating in the New Directions programwhile the other one hundred and fifty did not. The juveniles who didnot participate in the program acted as the control group in thisexperiment.
The300 juvenile offenders were then observed for between two months toeight months after they were placed on probation to see if they didor did not recidivate back into the criminal justice system.
Definitionof variables

Program
Thisis whether the juvenile offender in the sample participated in theNew Directions program

Offense
Thisis the type of offense that the juvenile committed to be put inprobation. The participants in this study include those who committedfelonies, status offences and misdemeanors.

Recidivate
Thisis whether juvenile in the study recidivated back into the criminaljustice system while on probation.

Age
Thisis the age of each juvenile offender in the study.

Race
Thisis the race of the juvenile offenders who participated in the study.It is classified into the black race and the white race.
Results
Fromthe analyses that were conducted this were the results
Achisquare test of independence was conducted to determine therelationship between participating in the New Directions program andwhether a juvenile offender recidivated or not. From the analysis,there exists a statistically significant relationship between the twovariables, X^{2}(1, N=300) =18.551, P<0. 000. Since the Cramer’s V=0.249, thisshows that the relationship is moderate.
Forthe juveniles who committed a felony offense, there was nosignificant relationship between participating in the offence andrecidivating,X^{2}(1, N=149) =2.799, P=0. 094. For the juveniles who committed amisdemeanor offense, there was a significant relationship betweenparticipating in the program and recidivating,X^{2}(1, N=126) =15.974, P<0. 000. This relationship is very strong(Cramer’s V=0.356). For the juveniles who committed a statusoffense, there was a significant relationship between participatingin the program and recidivating,X^{2}(1, N=25) =6.250, P=0. 012. The relationship is extremely strong(Cramer’s V=0.500).
Interms of age, for the juvenile offenders who were 13 years old, therewas no significant relationship between participating in the programand whether recidivating,X^{2}(1, N=64) =3.801, P=0.051. For the juvenile offenders who were 14years old, there was a significant relationship between participatingin the program and recidivating,X^{2}(1, N=62) =10.152, P=0. 001. This relationship is very strong(Cramer’s v=0.405). For the juvenile offenders who were 15 yearsold, there was a significant relationship between participating inthe program and recidivating,X^{2}(1, N=74) =7.466, P=0.006. This relationship is strong (Cramer’sv=0.318). For the juvenile offenders who were 16 years old, there wasa significant relationship between participating in the program andrecidivating,X^{2}(1, N=74) =9.615, P=0.002. This relationship is very strong (Cramer’sv=0.360). For the juvenile offenders who were 17 years old, therewas no significant relationship between participating in the programand recidivating,X^{2}(1, N=26) =1.354, P=0.245.
Interms of race, for the white juvenile offenders, there was nosignificant relationship between participating in the program andrecidivating,X^{2}(1, N=163) =1.125, P=0.289. For the black juvenile offenders, therewas a significant relationship between participating in the programand recidivating,X^{2}(1, N=137) =22.062, P<0. 000. This relationship is very strong(Cramer’s V=0.401).
Conclusions
Fromthe results above several conclusions can be reached. First,participating in the New Directions program results in lowerrecidivism rates. When we consider the type of offense the juvenilecommitted, participating in the new directions program was noteffective in reducing the recidivism rates. Participating in the NewDirections program was very effective in reducing recidivism rates injuveniles who committed misdemeanor and status offenses, but it ismore effective for the juveniles who committed status offenses.
Interms of age participating in the New Directions program was noteffective in reducing the recidivism rates among juveniles who wereaged 13 and 17. The New Directions program was effective in reducingthe recidivism rates among juveniles who were aged 14, 15 and 16 withit being more effective for the juveniles aged 14 years and leasteffective for juveniles aged 15 years.
Whenrace of the individuals was considered, the New Directions programwas not effective in reducing the recidivism rates for whitejuveniles, but it was very effective for the black juveniles.
Limitationsof the Research
Thisresearch study had one weakness. The sample size was quite small uponconsidering the number of juvenile offenders in the state. Havingthis kind of sample size might not be adequate to make conclusiveresults.
Appendix
Crosstabs
NEW OFFENSE – RECIDIVISM * PROGRAM ATTENDANCE Crosstabulation 

PROGRAM ATTENDANCE 
Total 

ATTENDED PROGRAM 
DID NOT ATTEND PROGR 

NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
103 
66 
169 

% within PROGRAM ATTENDANCE 
68.7% 
44.0% 
56.3% 

RECIDIVATED 
Count 
47 
84 
131 

% within PROGRAM ATTENDANCE 
31.3% 
56.0% 
43.7% 

Total 
Count 
150 
150 
300 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 
ChiSquare Tests 

Value 
df 
Asymp. Sig. (2sided) 
Exact Sig. (2sided) 
Exact Sig. (1sided) 

Pearson ChiSquare 
18.551^{a} 
1 
.000 

Continuity Correction^{b} 
17.562 
1 
.000 

Likelihood Ratio 
18.761 
1 
.000 

Fisher`s Exact Test 
.000 
.000 

LinearbyLinear Association 
18.489 
1 
.000 

N of Valid Cases 
300 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 65.50. 

b. Computed only for a 2×2 table 
Symmetric Measures 

Value 
Approx. Sig. 

Nominal by Nominal 
Phi 
.249 
.000 

Cramer`s V 
.249 
.000 

N of Valid Cases 
300 

a. Not assuming the null hypothesis. 

b. Using the asymptotic standard error assuming the null hypothesis. 
Crosstabs
ChiSquare Tests 

TYPE OF OFFENSE BEFORE PROBATION 
Value 
df 
Asymp. Sig. (2sided) 
Exact Sig. (2sided) 
Exact Sig. (1sided) 

FELONY 
Pearson ChiSquare 
2.799^{c} 
1 
.094 

Continuity Correction^{b} 
2.274 
1 
.132 

Likelihood Ratio 
2.806 
1 
.094 

Fisher`s Exact Test 
.102 
.066 

LinearbyLinear Association 
2.780 
1 
.095 

N of Valid Cases 
149 

MISDEMEANOR 
Pearson ChiSquare 
15.974^{d} 
1 
.000 

Continuity Correction^{b} 
14.518 
1 
.000 

Likelihood Ratio 
16.709 
1 
.000 

Fisher`s Exact Test 
.000 
.000 

LinearbyLinear Association 
15.847 
1 
.000 

N of Valid Cases 
126 

STATUS OFFENSE 
Pearson ChiSquare 
6.250^{e} 
1 
.012 

Continuity Correction^{b} 
4.340 
1 
.037 

Likelihood Ratio 
6.421 
1 
.011 

Fisher`s Exact Test 
.034 
.018 

LinearbyLinear Association 
6.000 
1 
.014 

N of Valid Cases 
25 

Total 
Pearson ChiSquare 
18.551^{a} 
1 
.000 

Continuity Correction^{b} 
17.562 
1 
.000 

Likelihood Ratio 
18.761 
1 
.000 

Fisher`s Exact Test 
.000 
.000 

LinearbyLinear Association 
18.489 
1 
.000 

N of Valid Cases 
300 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 65.50. 

b. Computed only for a 2×2 table 

c. 0 cells (.0%) have expected count less than 5. The minimum expected count is 31.93. 

d. 0 cells (.0%) have expected count less than 5. The minimum expected count is 20.71. 

e. 1 cells (25.0%) have expected count less than 5. The minimum expected count is 4.00. 
Symmetric Measures 

TYPE OF OFFENSE BEFORE PROBATION 
Value 
Approx. Sig. 

FELONY 
Nominal by Nominal 
Phi 
.137 
.094 

Cramer`s V 
.137 
.094 

N of Valid Cases 
149 

MISDEMEANOR 
Nominal by Nominal 
Phi 
.356 
.000 

Cramer`s V 
.356 
.000 

N of Valid Cases 
126 

STATUS OFFENSE 
Nominal by Nominal 
Phi 
.500 
.012 

Cramer`s V 
.500 
.012 

N of Valid Cases 
25 

Total 
Nominal by Nominal 
Phi 
.249 
.000 

Cramer`s V 
.249 
.000 

N of Valid Cases 
300 

a. Not assuming the null hypothesis. 

b. Using the asymptotic standard error assuming the null hypothesis. 
Crosstabs
NEW OFFENSE – RECIDIVISM * PROGRAM ATTENDANCE * AGE OF JUVENILE ON PROBATION Crosstabulation 

AGE OF JUVENILE ON PROBATION 
PROGRAM ATTENDANCE 
Total 

ATTENDED PROGRAM 
DID NOT ATTEND PROGR 

13 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
20 
24 
44 
% within PROGRAM ATTENDANCE 
83.3% 
60.0% 
68.8% 

RECIDIVATED 
Count 
4 
16 
20 

% within PROGRAM ATTENDANCE 
16.7% 
40.0% 
31.2% 

Total 
Count 
24 
40 
64 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

14 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
20 
18 
38 
% within PROGRAM ATTENDANCE 
87.0% 
46.2% 
61.3% 

RECIDIVATED 
Count 
3 
21 
24 

% within PROGRAM ATTENDANCE 
13.0% 
53.8% 
38.7% 

Total 
Count 
23 
39 
62 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

15 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
38 
19 
57 
% within PROGRAM ATTENDANCE 
88.4% 
61.3% 
77.0% 

RECIDIVATED 
Count 
5 
12 
17 

% within PROGRAM ATTENDANCE 
11.6% 
38.7% 
23.0% 

Total 
Count 
43 
31 
74 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

16 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
23 
5 
28 
% within PROGRAM ATTENDANCE 
52.3% 
16.7% 
37.8% 

RECIDIVATED 
Count 
21 
25 
46 

% within PROGRAM ATTENDANCE 
47.7% 
83.3% 
62.2% 

Total 
Count 
44 
30 
74 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

17 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
2 
0 
2 
% within PROGRAM ATTENDANCE 
12.5% 
0.0% 
7.7% 

RECIDIVATED 
Count 
14 
10 
24 

% within PROGRAM ATTENDANCE 
87.5% 
100.0% 
92.3% 

Total 
Count 
16 
10 
26 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

Total 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
103 
66 
169 
% within PROGRAM ATTENDANCE 
68.7% 
44.0% 
56.3% 

RECIDIVATED 
Count 
47 
84 
131 

% within PROGRAM ATTENDANCE 
31.3% 
56.0% 
43.7% 

Total 
Count 
150 
150 
300 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 
ChiSquare Tests 

AGE OF JUVENILE ON PROBATION 
Value 
df 
Asymp. Sig. (2sided) 
Exact Sig. (2sided) 
Exact Sig. (1sided) 

13 
Pearson ChiSquare 
3.801^{c} 
1 
.051 

Continuity Correction^{b} 
2.793 
1 
.095 

Likelihood Ratio 
4.031 
1 
.045 

Fisher`s Exact Test 
.058 
.045 

LinearbyLinear Association 
3.742 
1 
.053 

N of Valid Cases 
64 

14 
Pearson ChiSquare 
10.152^{d} 
1 
.001 

Continuity Correction^{b} 
8.505 
1 
.004 

Likelihood Ratio 
11.115 
1 
.001 

Fisher`s Exact Test 
.003 
.001 

LinearbyLinear Association 
9.989 
1 
.002 

N of Valid Cases 
62 

15 
Pearson ChiSquare 
7.466^{e} 
1 
.006 

Continuity Correction^{b} 
6.014 
1 
.014 

Likelihood Ratio 
7.471 
1 
.006 

Fisher`s Exact Test 
.011 
.007 

LinearbyLinear Association 
7.365 
1 
.007 

N of Valid Cases 
74 

16 
Pearson ChiSquare 
9.615^{f} 
1 
.002 

Continuity Correction^{b} 
8.161 
1 
.004 

Likelihood Ratio 
10.223 
1 
.001 

Fisher`s Exact Test 
.003 
.002 

LinearbyLinear Association 
9.485 
1 
.002 

N of Valid Cases 
74 

17 
Pearson ChiSquare 
1.354^{g} 
1 
.245 

Continuity Correction^{b} 
.166 
1 
.684 

Likelihood Ratio 
2.045 
1 
.153 

Fisher`s Exact Test 
.508 
.369 

LinearbyLinear Association 
1.302 
1 
.254 

N of Valid Cases 
26 

Total 
Pearson ChiSquare 
18.551^{a} 
1 
.000 

Continuity Correction^{b} 
17.562 
1 
.000 

Likelihood Ratio 
18.761 
1 
.000 

Fisher`s Exact Test 
.000 
.000 

LinearbyLinear Association 
18.489 
1 
.000 

N of Valid Cases 
300 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 65.50. 

b. Computed only for a 2×2 table 

c. 0 cells (.0%) have expected count less than 5. The minimum expected count is 7.50. 

d. 0 cells (.0%) have expected count less than 5. The minimum expected count is 8.90. 

e. 0 cells (.0%) have expected count less than 5. The minimum expected count is 7.12. 

f. 0 cells (.0%) have expected count less than 5. The minimum expected count is 11.35. 

g. 2 cells (50.0%) have expected count less than 5. The minimum expected count is .77. 
Symmetric Measures 

AGE OF JUVENILE ON PROBATION 
Value 
Approx. Sig. 

13 
Nominal by Nominal 
Phi 
.244 
.051 
Cramer`s V 
.244 
.051 

N of Valid Cases 
64 

14 
Nominal by Nominal 
Phi 
.405 
.001 
Cramer`s V 
.405 
.001 

N of Valid Cases 
62 

15 
Nominal by Nominal 
Phi 
.318 
.006 
Cramer`s V 
.318 
.006 

N of Valid Cases 
74 

16 
Nominal by Nominal 
Phi 
.360 
.002 
Cramer`s V 
.360 
.002 

N of Valid Cases 
74 

17 
Nominal by Nominal 
Phi 
.228 
.245 
Cramer`s V 
.228 
.245 

N of Valid Cases 
26 

Total 
Nominal by Nominal 
Phi 
.249 
.000 
Cramer`s V 
.249 
.000 

N of Valid Cases 
300 

a. Not assuming the null hypothesis. 

b. Using the asymptotic standard error assuming the null hypothesis. 
Crosstabs
NEW OFFENSE – RECIDIVISM * PROGRAM ATTENDANCE * RACE OF JUVENILE Crosstabulation 

RACE OF JUVENILE 
PROGRAM ATTENDANCE 
Total 

ATTENDED PROGRAM 
DID NOT ATTEND PROGR 

WHITE 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
62 
41 
103 

% within PROGRAM ATTENDANCE 
66.7% 
58.6% 
63.2% 

RECIDIVATED 
Count 
31 
29 
60 

% within PROGRAM ATTENDANCE 
33.3% 
41.4% 
36.8% 

Total 
Count 
93 
70 
163 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

BLACK 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
41 
25 
66 

% within PROGRAM ATTENDANCE 
71.9% 
31.2% 
48.2% 

RECIDIVATED 
Count 
16 
55 
71 

% within PROGRAM ATTENDANCE 
28.1% 
68.8% 
51.8% 

Total 
Count 
57 
80 
137 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 

Total 
NEW OFFENSE – RECIDIVISM 
NONRECIDIVATED 
Count 
103 
66 
169 

% within PROGRAM ATTENDANCE 
68.7% 
44.0% 
56.3% 

RECIDIVATED 
Count 
47 
84 
131 

% within PROGRAM ATTENDANCE 
31.3% 
56.0% 
43.7% 

Total 
Count 
150 
150 
300 

% within PROGRAM ATTENDANCE 
100.0% 
100.0% 
100.0% 
ChiSquare Tests 

RACE OF JUVENILE 
Value 
df 
Asymp. Sig. (2sided) 
Exact Sig. (2sided) 
Exact Sig. (1sided) 

WHITE 
Pearson ChiSquare 
1.125^{c} 
1 
.289 

Continuity Correction^{b} 
.804 
1 
.370 

Likelihood Ratio 
1.122 
1 
.289 

Fisher`s Exact Test 
.327 
.185 

LinearbyLinear Association 
1.118 
1 
.290 

N of Valid Cases 
163 

BLACK 
Pearson ChiSquare 
22.062^{d} 
1 
.000 

Continuity Correction^{b} 
20.463 
1 
.000 

Likelihood Ratio 
22.694 
1 
.000 

Fisher`s Exact Test 
.000 
.000 

LinearbyLinear Association 
21.901 
1 
.000 

N of Valid Cases 
137 

Total 
Pearson ChiSquare 
18.551^{a} 
1 
.000 

Continuity Correction^{b} 
17.562 
1 
.000 

Likelihood Ratio 
18.761 
1 
.000 

Fisher`s Exact Test 
.000 
.000 

LinearbyLinear Association 
18.489 
1 
.000 

N of Valid Cases 
300 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 65.50. 

b. Computed only for a 2×2 table 

c. 0 cells (.0%) have expected count less than 5. The minimum expected count is 25.77. 

d. 0 cells (.0%) have expected count less than 5. The minimum expected count is 27.46. 
Symmetric Measures 

RACE OF JUVENILE 
Value 
Approx. Sig. 

WHITE 
Nominal by Nominal 
Phi 
.083 
.289 
Cramer`s V 
.083 
.289 

N of Valid Cases 
163 

BLACK 
Nominal by Nominal 
Phi 
.401 
.000 
Cramer`s V 
.401 
.000 

N of Valid Cases 
137 

Total 
Nominal by Nominal 
Phi 
.249 
.000 
Cramer`s V 
.249 
.000 

N of Valid Cases 
300 

a. Not assuming the null hypothesis. 

b. Using the asymptotic standard error assuming the null hypothesis. 