Attachment J-2 Beneficiary Satisfaction Survey

SECTION J ATTACHMENT J-2

BENEFICIARY SATISFACTION SURVEY METHODOLOGY

Survey Methodology:
General: Telephone surveys will be considered complete with responses to the following question (or a question of a similar nature): All things considered, how satisfied were you with their processing of your claim, their customer service or any other service you received from (name of contractor) during the past 3 months?
While follow on questions may be asked during the survey, the incentive will be based solely on this question (or a similar question).
Methodology: Responses will be scored on 7-point scale with the following meaning:

1 = completely dissatisfied

2 = very dissatisfied

3 = somewhat dissatisfied

4 = neither satisfied nor dissatisfied

5 = somewhat satisfied

6 = very satisfied

7 = completely satisfied.
The sample for this survey will be drawn from a list of outpatient purchased care visits developed by DHA from TRICARE Encounter Data (TED). Telephone numbers will be obtained by the survey firm from commercial sources and/or for contact with a randomly sampled population of users.
Highly satisfied” with purchased care services will be defined as a survey response of 6 (Very Satisfied) or 7 (Completely Satisfied) on the 7-point satisfaction scale. “Percent Highly Satisfied” will be calculated by dividing the number of respondents who select a rating of 6 or 7 by total number of respondents who provide a valid response to the applicable satisfaction question.
Estimation of Sample Size: The target response pool of 402 was developed with an interest to achieve an estimate of "percent satisfied" to three percentage points on a onehundred-point scale with 95 percent confidence. It is assumed that approx. 90% of users will be satisfied with the service(s) and that the distribution of responses to the overall satisfaction item will be binomial (i.e. satisfied, not satisfied). A binomial distribution with p = 0.9, a 95% confidence interval with half length .03 (i.e. +/- 3%) requires a sample of about 400.
The calculation is: using the normal approximation to binomial distribution, .03=2*SD & SD--sqrt(p(1-p)/n)=sqrt(.9*.1/n). This means that .015=sqrt(.9*.1/n), so sqrt(n)=.3/.015=20, or n=400.
Minimal Surveys Required: Telephone survey efforts will continue until 400 completions are accomplished for the services received in the quarter. The telephone survey process will not be considered final until attaining 400 completed responses for the period surveyed.