Supporting Statement B

B. Collection of Information Employing Statistical Methods

1. Respondent Universe and Sampling Methods

The data collection will use statistical methods to select respondents. Five percent (5%) (n=583) of subscribers to the platform will have four weeks to voluntarily participate in the web-based survey. The responses are intended to serve as feedback to help CDC/PHLP and the individual platform meet subscriber needs.

The potential respondent universe comprises over 11,658 public health legal professionals. The universe contains subscribers – as determined from News subscriber email address domains – from different levels of government (.gov, .us, or .mil)(21.09%), universities (.edu)(10.56%), non-profit organizations (.org)(11.67%), international subscribers (represented by email addresses indicating country of origin)(4.75%), and a category representing individuals, private businesses, or corporations (.com and .net)(51.85%). The remaining domains (.coop, .jobs, .local, .info, .biz) comprise .08% of subscribers. Survey participants will be randomly chosen from among the 6 different strata described here.

Sector by Email Domain	% of Total Population (11658)	Total Subscribers per Sector	5% of Subscribers per Sector
individuals, businesses (.com / .net)	51.85	6045	302.25
government (.gov / .us / .mil)	21.09	2459	122.95
non-profit organizations (.org)	11.67	1360	68
education (.edu)	10.56	1231	61.55
international (.int)	4.75	554	27.7
other (.coop / .jobs / .local / .info / .biz)	.08	9	0.45
TOTAL	100	11,658	582.9

We estimate an approximately 50% response rate for News subscribers asked to participate in the survey. This figure is based on a review of literature regarding web-based surveys.

2. Procedures for the Collection of Information

Readers subscribe to the News by submitting their email address. No information is known about subscribers’ sector or field of employment except for what may be known from the subscribing email address domain names. Here, we have categorized subscribers based on work sector (e.g., government, education, non-profit) as indicated by email address domain. Next, the proportion of each of the five categories in the entire subscriber universe was calculated. Finally, the number of subscribers to be asked to participate in the survey was calculated by finding 5% of each of the five categories. Five percent of survey respondents will be chosen at random from each of the five categories.

This survey is intended to establish a baseline for customer satisfaction. The data collected will be stratified in different ways during the analysis. We plan to examine responses primarily by profession, sector, and use of the platform. Most of the survey questions are not in a “yes/no” format. For many of the questions, respondents are asked to select their preferences, and at times are given the option of more than 5 possible responses. The data collection is expected to have relationships with other data sets – e.g. we intend to compare usage by .edu against .gov users – in order to determine the utility of the product based on employment sector.

The degree of accuracy needed is the standard 95% confidence level due to the nature of the survey, which includes mostly multiple-choice.

No unusual problems are expected that would require specialized sampling procedures.

The online instrument will be given to the representative sample. Respondents will be given advanced notice of the timeline to return the survey.

3. Methods to Maximize Response Rates and Deal with Nonresponse

In order to avoid a response rate of less than 50% and maximize responses, a number of measures have been implemented:

• The pool of potential respondents only includes current subscribers to the platform and therefore the survey is directly relevant;

• The survey will be pre-announced in an issue of the CDC Public Health Law News prior to the mailing of individual invitations

• The invitation to the survey (Attachment 4) will be sent to each chosen subscriber’s personal email address, greatly improving the probability that the invitation will be viewed and acted upon;

• The invitation will include a brief explanation of the survey and the benefit of the subscriber’s feedback for the platform and CDC;

• Individual readers who have been invited to participate in the survey will receive individualized letters prompting them to complete the survey if they do not do so by the end of week 1 and again at the end of week 3. (Attachment 3)

The survey has been designed to minimize respondent burden. The survey is logical, easy to read, and relatively quick to complete. The survey is also web-based, which will allow for a visually appealing layout.

4. Tests of Procedures or Methods to be Undertaken

The survey instrument has been pilot tested with 9 or fewer respondents. We have performed technical testing to ensure technical quality, ease of use, and proper collection of data. We have tested the questions with fewer than 9 people from CDC to ensure questions and answers are logical and clear.

5. Individuals Consulted on Statistical Aspects and Individuals Collecting and/or Analyzing Data

Data collection and analysis will be performed by PHLP staff. We anticipate the data to be straightforward and not require extremely complex analysis techniques.

CDC/PHLP staff:

Rachel Weiss, J.D.

Public Health Analyst

CDC Public Health Law Program

Phone: 404.639.5031

Email: elo3@cdc.gov

Kevin Cartwright

Northrop Grumman Corp.

Web Developer contracted to:

CDC Public Health Law Program

Phone: 404.639.5032

Email: krc4@cdc.gov