Defense Acquisition Research Journal Issue 95

January 2021

Hotelling's T-Squared Distribution in Statistics The research questions we are attempting to answer in this section are: • When all the survey responses for each subgroup are taken together as a whole, are there statistical differences in the responses?

° Are the perceptions of the Direct-Control system differ ent when Less Info is provided, More Info is available, or a hands-on experiment is conducted? ° Are the perceptions of the Remote-Control system differ ent when Less Info is provided, More Info is available, or a hands-on experiment is conducted? ° Are the perceptions of the Autonomous system different when Less Info is provided, More Info is available, or a hands-on experiment is conducted?

Tables 3 and 4 show a sampling of the results from the Hotelling T 2 test. The null hypothesis is that no statistical differences result from using a parametric Hotelling T 2 test, where all of the survey responses in each of the subcategories, when taken together, simultaneously do not indicate that any differences are discernible between the two groups tested.

TABLE 3. HOTELLING Hotelling Test Groups

P -value

Variables Tested

Less Info vs. More Info

0.5863

VAR1:VAR14 vs. VAR101:VAR114

Less Info vs. More Info

0.7998

VAR15:VAR25 vs. VAR115:VAR125

Less Info vs. More Info

0.3515

VAR26:VAR36 vs. VAR126:VAR136

Less Info vs. More Info

0.2084

VAR37:VAR47 vs. VAR137:VAR147

Less Info vs. More Info

0.7095

VAR48:VAR51 vs. VAR148:VAR151

Less Info vs. More Info

0.4475

VAR52:VAR54 vs. VAR152:VAR154

Less Info vs. Experiment

0.0000

VAR15:VAR25 vs. VAR415:VAR425

Less Info vs. Experiment

0.0144

VAR26:VAR36 vs. VAR426:VAR436

Less Info vs. Experiment

0.0793

VAR37:VAR47 vs. VAR437:VAR447

More Info vs. Experiment

0.0000

VAR115:VAR125 vs. VAR415:VAR425

More Info vs. Experiment

0.1215

VAR126:VAR136 vs. VAR426:VAR436

More Info vs. Experiment

0.3232

VAR137:VAR147 vs. VAR437:VAR447

15

Defense ARJ, January 2021, Vol. 28 No. 1 : 2-39