Ver-A-Fast is one of the largest verification companies in the world, making over 15 million calls annually. Since 1976, Ver-A-Fast has been a pioneer in the field of customer relations and data collection. We currently work with over 150 newspapers across the country and conduct over 550 usage studies/audits annually. Some of our clients include:

Ver-A-Fast's success is based on the value of short, precise surveys. We operate on the premise that there can be a direct relationship between length of a questionnaire and the accuracy of data collected.

The longer a person is asked questions about a product or service, the greater the probability that the answers being recorded will become tainted or inaccurate. A person may begin to answer questions in a way that they perceive will end the survey in the quickest possible manner. This may not necessarily be how they truly feel about the product or service being studied. Also, is the person that willfully ends up answering a lengthy questionnaire (taking 20 or 30 minutes) a representative sample of the population being measured?

Our surveys are designed to take less than 3 minutes; asking only the most pertinent questions. If more information is desired, a second study is recommended. This keeps the reported information clear and easy to understand.

Ver-A-Fast's typical usage study is comprised of completed surveys, incomplete surveys, refusals, and inaccurate numbers.

• A completed survey is any adult individual that answers all questions up to an "End Survey" response.
• An incomplete survey is any adult individual that terminates the survey before reaching an "End Survey" response. Due to the fact that the questionnaire was not finished, data from questions answered are dismissed.
• A refusal is any adult individual that does not agree to answer questions.
• An inaccurate number is a wrong number, a new phone number, a disconnected number, or a number that has been changed to a non-published number.

Each question on our survey is designed to extract specific information from the caller. Sentence structure and arrangement has been designed with the aid of Dr. James Martin, Professor of Marketing and Management, John Carroll University.

TO GENERATE A SAMPLE in the most accurate and cost-effective manner, the sample is comprised of calls made randomly throughout the white pages of the telephone directory being studied. When the white pages are received electronically, the file is sorted by the 7th digit of the phone number (i.e. 440-331-0250) to provide a random sample. These processes provide a statistically valid measure of the book's total distribution area.

TIMING of the survey is normally 45 to 90 days after distribution has been completed. If a competitor's book is delivered during that time, an extra 30 days is added before beginning the study. This provides ample cool-down time to eliminate a bias in favor of the book most recently distributed from affecting the results of the study.

OUR DATA IS REPORTED back to the publisher in an easy-to-read, understandable format which includes a master report, color graphs, and the actual survey forms used by our operators. The survey forms list the name and address of the person spoken to, telephone number, and gender as well as the date contacted and the Ver-A-Fast operator number. The survey forms are mailed to the publisher so that they can be referenced to substantiate the results of the survey.

QUALITY is ensured through our multi-step approach:

• Several operators are used per project. This allows us to compare results and detect any inaccuracies. Any variance among operators of greater than 10% is automatically researched further.
• 10% of all calls made are recalled to verify that the operator asked all necessary questions and reported the information correctly.
• Calls are cross-referenced against a daily download of all Ver-A-Fast telephone activity from our primary long distance carrier. Telephone numbers and call durations are checked against our operator's work to ensure that the calls were made.
  
  

THE MATHEMATICALLY ACCEPTED FORMULA to decide the number of surveys to be completed is outlined below (where Z=1.96, p=.5, q=.5, and B=.05):

	B	The precision desired in absolute percentage points.
	Z	Standard score used to represent the confidence level desired. To be 95% confident about the sample's precision, you use a Z score of 1.96. To be 99% confident about the sample's precision, you use a Z score of 2.58. To be 99.7% confident about the sample's precision, you use a Z score of 3.
		The standard deviation for a proportion, where p is the proportion of incidence in the population and q=1-p. This represents how different the people are in the population. Using p=.5 and q=.5, you ar using the safest, or most conservative values, as these will give you the highest standard deviation possible. This ensures your sample size will be large enough to avoid significant sampling errors.
	Note: This formula is recommended for use by the Advertising Research Foundation in the ARF GUIDELINES FOR THE MEASUREMENT OF THE YELLOW PAGES MEDIUM.

This formula states that 385 completed surveys yields a 95% confidence that the sample drawn is within "5 percentage points of the population proportion - assuming that the population is homogeneous and greater than 2500 households (listings). A homogeneous population means that every household (listing) in the population has the same likelihood of being surveyed.