CRIM 320 Lecture Notes - Explained Variation, Grater, John Tukey
17 views4 pages
For unlimited access to Class Notes, a Class+ subscription is required.
Difference between more than two population means
-Also called single-factor ANOVA
-Statistical test for comparing the sample means of three or more INDEPENDENT groups in
terms of a continuous dependent measure
ANOVA versus t-test
-T-test can only compare two samples at the same time.
-With t-test are tested at .05, 1 in 20 chance of falsely rejecting the null hypothesis.
-Therefore, with 21 t-test, one should be significant by chance alone.
THREE OR MORE GROUPS
One nominal or ordinal variable with more than two attributes considered as the IV
-onset of delinquency:
1)childhood-onset (before 12)
2)adolescent-onset (between 13-17)
3) adult-onset offenders (18 and older)
-Type of offenders:
1)first-time offenders (no prior record)
2)recidivists (only 1 prior conviction)
3) multi-recidivists (2 or more prior convictions)
-Police interventions for a call of partner abuse:
3)ask the husband to leave the house for 24 hours
-One interval or ratio variable considered
-persistence of criminal activity
-time spend without committing a crime after release from prison
EXAMPLE OF RESEARCH QUESTIONS
-H(1): Age of onset of criminal activity has an impact on the persistence of criminal activity
-H(2): Type of offenders differ as to the time spent without committing a crime after being
released from prison
-H(3): Type of police interventions influence the number of subsequent police contact for
ANALYSIS OF VARIANCE
-Does not look at the mean directly but at the variance