Part B 3060-0647 (April 2009)

B. Collections of Information Employing Statistical Methods:

1. There is a potential respondent universe of approximately 33,000 cable community units. This survey will be sent to the cable operators serving a randomly selected sample of these cable community units to obtain information at the community level. To determine the number of observations needed for statistical precision in our samples, we applied a standard statistical formula. Sample-size formulas require the user to choose a limit in terms of how much it is expected that the sample results might randomly fluctuate from the actual averages of the prices being measured. Larger sample sizes lead to less fluctuation but also increase the cost of the survey. We chose a sample size that strikes a reasonable balance between benefit and cost. We limited the allowable error to within 50 cents of actual price with a 95 percent probability. In other words, if we were to take repeated random samples, the results would fall within 50 cents of the actual values in 95 out of 100 samples.

Taking these factors into account along with the statistical variance found in previous surveys, we applied a standard statistical formula in order to calculate the size of the sample that would be required to maintain the desired statistical confidence level. We applied this formula to two groups of cable operators: (1) those subject to effective competition, as determined by the Commission, in accordance with the statutory definitions of effective competition; and (2) those not subject to effective competition, which make up the balance of cable operators.

Before selecting the sample for operators facing effective competition, in order to make sure that our sample is representative, we stratified that group by the test used to determine effective competition and selected a portion of the sample from each stratum. We stratified the non-effective-competition group by separating it into small, medium, and large operators (in terms of number of subscribers) and selected a portion of the sample from each stratum. For both the effective competition and the non-effective-competition groups, larger operators had a higher chance of being selected than smaller operators. This not only made our sample more representative of all cable subscribers, but also reduced the burden on smaller cable operators.

2. Because this is a mandatory information collection, the Commission expects a very high response rate. Based on the response rate to previous surveys, we estimate that the response rate to the main Survey will be 96%.

3. We make the questionnaire available to respondents via the Commission's Internet site and receive completed questionnaires from respondents via the Internet. In order to obtain the highest possible response rate with useable data, we make follow-up telephone calls to all recipients who fail to respond in a timely manner and to respondents who provide incomplete data or data with apparent inaccuracies.

4. The Commission has built on experience gained in conducting previous surveys of this type. After each survey, we review the consistency of responses in terms of apparent understanding of the questions and adjust future questions accordingly.

5. John Scott, industry economist in the Media Bureau, designed this and previous surveys of cable industry prices. He is responsible for analyzing the data received from respondents. He can be reached at (202) 418-2330.