METHODOLOGICAL NOTE

While targeted lists can help increase data collection efficienccy, it's very important that crucial aspects of the list's composition be fully understood.

For many studies, the screening required to reach a sufficient number of households with specific characteristics often involve costs that are beyond the client's means, or are simply not justified by the study's objectives. There are, of course, listed household sampling options available for targeting households with family members of specific age/gender or household income. In addition, there are list options that can be used for targeting households with specific product ownership/usage, purchase behavior, interests, hobbies,and other household characteristics.

There are literally hundreds of lists that can be used to generate a targeted sample.

Any time a list is compiled, it is a potential sample source. For example, a list from personal computer warranty cards is compiled into a database for registration and repair purposes. This list can also be used as a source of sample for surveying owners of PCs or related products and services. Subscription lists,trade show attendance lists, a list of those who responded to a promotion campaign, etc., are all examples of targeted
list possibilities.

Since there are so many possible targeted list sources, it would be futile for us to attempt  to review all options. Rather, this Methodological Note reviews some global issues relating to targeted list options and examines some common problems and limitations encountered when using targeted lists. It concludes with information regarding one ofthe most popular, multipurpose targeted list sources.

E fficiency Versus Representation 

The primary reason to use a targeted list for sampling purposes is to increase data collection efficiencies (i.e., keep costs down). Unfortunately,this often means you have to make compromises on the coverage of your target market (representation). Typically, targeted lists do not represent all of the target market, just those who happen to have gotten on the list.

Once you've made the decision to utilize a targeted list (instead of an RDD approach),your preliminary investigations will most likely lead to numerous possible list sources. As with any list of this type, it is imperative that you examine the way in which the list is compiled and maintained. Regardless of whether you are dealing with a sampling company, a list broker or with the actual compiler of the list, you will need to determine exactly what each option means in order to determine which list, if any, will meet the objectives of the study. On the surface, a list may appear ideal for your application,
but upon closer study, and for a variety of reasons, it may in fact not be satisfactory. Make sure you find out where the information comes from, how
often it's updated, what record information is available, etc.

GENESYS Comments:

Geographic Representation. Since lists are compiled for a variety of reasons, they may reach the particular characteristic you're interested in but have significant compromises on geographic coverage. For example, in searching for a list source for woman golfers, we came across two options; a golf magazine subscription list and an attendee list from a golf trade show. On the surface both options seemed acceptable, in terms of efficiently locating women golfers. Upon further investigation, we discovered that the trade show was held in only one large city, and the attendees were local. The subscribers to the golf magazine were from all over the country. Depending on your objective, these two options can provide different results.

Efficiency -Beware. On the surface, a targeted list may appear to be ideal, but may actually result in the high data collection costs that you were trying to avoid in the first place. Every record on a list has to meet some criteria or qualification in order to get on the list. However, that doesn't necessarily mean that, over time, a record will continue to meet those same qualifications. For example, some targeted lists continually add new records, but
only make minimal efforts to remove obsolete records (i.e., when the household has moved). Further, if a list is compiled based on a characteristic
like product ownership, some of the records will eventually be for households that no longer own the product (regardless of whether the household has moved or not). In both of these situations, the net result is higher data collection costs.

How can these problems be avoided? The basic list evaluation questions will, for the most part, prevent these types of situations. (See the related GENESYS Methodological Note, "How to Evaluate List Sources").

List Availability

While there are many list options,not all compiled lists are available for every survey research application due to restrictions placed on them by the compilers. Some lists may not be available for researchers at all, others will only be released with stipulations restricting their use. You may also be required to provide a copy of the survey instrument. Another common situation is that a list will be available, but with restrictions on what pieces of record information you can obtain. Serious difficulties can arise if you're doing a phone survey and the supplier won't release the phone numbers.

GENESYS Comments:

You have a right to know. Don't be surprised if you encounter some resistance when you ask questions about a particular list. For reasons we've never understood, list compilers, list brokers, and even sampling companies are often unwilling (or unable) to answer some very basic questions about their lists. This situation,when encountered, raises some doubt in our minds as to the value ofthe list. When we investigate a list source, our objective is to
find out exactly what the list represents so that you can determine if this source meets your objectives. It's our job to provide you with all the options.

Providing Proof. You may be requested to provide a copy of your survey document. Complicating matters is the fact that sometimes they won't even begin processing your order until you provide the document. This can be problematic. If your questionnaire is still being written, but won't be ready for a week, and the sample needs to be ordered immediately (so that it can be processed in time for when interviewing begins in a week) you have a timing problem!

This is usually not a serious difficulty. We have found that just providing a screener, or the first few pages of a preliminary questionnaire will often suffice.

Survey Versus Sales. Some lists will not be released for telemarketing purposes, but are available for survey research applications. Don't be surprised if at the onset, the person on the other end of the phone is a little fuzzy on the difference and doesn't want to release the list. As we're sure you've encountered, there are those in the outside world who do not recognize that the research industry does not sell anything. We've found that after providing
a little background, education, clarification and assurance (that the list will be used for survey purposes only), the list will be released.

Multipurpose Lists

There are literally hundreds of categories that can be used for sample targeting. One such list is complied by Donnelley Marketing and is known as their Shareforce List. The Shareforce List is compiled from responses to a mail questionnaire that is dropped 10 times a year. The volume of returns varies, but it ranges from 1 to 2 million (this represents, approximately, a 5% response rate.) The questionnaire includes questions about household purchases,
purchase intent, interests, and a wide variety ofother lifestyle issues.

The response categories are varied. Some examples from recent questionnaires provide a good idea of the nature of the categories:

• Interested in Fishing
• Interested in Flying
• Owns a PC -IBM/Apple
• Female Business Owner
• Have a Cellular Phone
• Federal Worker
• Buys Videotapes
• Working Couple
• Planing to Buy a Vehicle/Car/Pickup/etc.

As with all list sources, the Shareforce List also has some inherent limitations that need to be addressed in terms of study objectives. For example:

• Some of the questions asked in a particular Shareforce questionnaire, may remain in subsequent mailings for a number of years,others may only be included once,or every few years. This means that the information's currency must be addressed. For example, questions regarding personal computer ownership have been inserted and removed numerous times. Just because a household had a particular PC brand two years ago does not necessarily mean they still have it.
• Self-administration of the questionnaire and how a question is worded can have significant impacts on what the resultant targeted list really represents. For example, the Shareforce questionnaire frequently asks the respondent to choose items of interest and hobbies from a list. One of the items on the list is "traver'. However, this does not indicate that this respondent has actually traveled - they may only have an interest in travel.

It should also be noted that Shareforce is widely marketed by list brokers and sampling companies under other names. We have even seen promotions where the name was changed, the exact Shareforce categories were listed (word for word),and the product was presented as an exclusive, proprietary source compiled by the broker themselves. Unfortunately, there are no standards for disclosure in either the list brokerage or sampling business.

As with all the sampling products provided by GENESYS Sampling Systems, full disclosure of sources, methodologies, etc., is standard procedure. After all, this is the only way for you to effectively evaluate alternative sources/options for meeting your client's objectives.

If you have a question as to whether the Shareforce List (or any other targeted lists) can be of use in your project,we will be more than happy to discuss with you the pros and cons of each alternative.