## E. Computerized SAC and SOAC Versions and Critical Difference (CDs)

We recommend use of the 2007 computerized versions of SAC/SOAC whenever possible for pre and post testing (outcome measures).. The standard deviations (SDs) and test-retest correlations (r) from our data set of 60 were used to calculate standard error of measurement (SEM) and critical differences (CD). CD's were found to be 5.2 (14.4%) for SAC and 4.0 (11.2%) for SOAC. The CDs we have for paper pencil versions of SAC/SOAC are from the original study by Short. They are 23% for SAC and 16% for SOAC. These 1982 paper pencil CDs are slightly larger partly because the test-retest correlations were not as good. This is typical for other CD numbers to be higher on paper and pencil versions as reported by Holcomb and Punch (2006) and Weinstein, Spitzer, and Ventry (1986). If the CD number or % is subtracted from the unaided pre-test score, any number or % smaller than this result, on the outcome score is considered to be significant at the 95% confidence interval for true change. For example, if the computerized SAC pre-test equals 50%, and the post-test outcome score = 35% this is a significant difference.

Correlations (r) |
Critical Differences |
|||||

SAC |
SOAC |
SAC |
SOAC |
|||

Num |
% |
Num |
% |
|||

Pencil/Paper |
.80 |
.90 |
9.0 |
23.0 |
6.5 |
16.0 |

Computerized |
.94 |
.97 |
5.2 |
14.4 |
4.0 |
11.2 |

We recommend use of the following scale for interpretation of percent scores of either paper-pencil or computerized versions (both 1982 and 2007 versions).

0-20 % |
No disability or handicap |

21-40 % |
Slight hearing disability or handicap |

41-70 % |
Mild to moderate disability or handicap |

71-100 % |
Severe hearing disability or handicap |

To convert raw scores to percent, divide the raw score total by the number of items (such as 9 items in the total or 5 items for the disability section). Then, subtract 1 and multiply x 25 = %.