Login - CS Members

We have more than 15 years of experience examining residential technologies. Our market analyses combine primary consumer research with industry knowledge, giving our clients a comprehensive demand- and supply-side view of emerging markets.

Our research covers the entire value chain and the multiple intersections where computing touches entertainment, home control, security, communications, and consumer avocations.

What's the typical design for custom research?

Why bother with focus groups?

How do you gather quantitative data?

What is a national sample?

Why do you have to oversample to provide results for a specific segment of the population?

Why not sample even more households?

How can I ever know that the consumer research I receive will hold up?

Developing a questionnaire doesn't seem that hard. What's the big deal?

• An Executive Summary of existing research compilation from Parks Associates’ consumer and industry research.
• Qualitative research in the form of:
  • focus groups; and/or
  • one-on-one interviews.
• Quantitative Research with Appropriate Population Sets

Parks Associates seeks to provide its clients with the most meaningful research at its disposal while also keeping an eye on cost effectiveness. Thus, we typically propose compiling both our own and available consumer and industry research as a precursor to custom consumer research or as an adjunct to help manage costs.

^ top

Parks Associates recommends hosting focus groups or one-on-one research before fielding large national studies for a variety of reasons. Focus groups cannot, of course, generalize to the total population, but they have significant value, well beyond even their dollar cost, when completed in a manner that optimizes their research output potential. When competently completed, focus groups provide:

• Target segments’ attitudes, general understanding, use of language, and even objections towards the areas under investigation.
• Moreover, focus groups allow mapping of the consumer thinking process.
• Focus groups manage to allow both spontaneity and discussion among participants as well as mapping development of the underlying thinking structure about the topic under investigation.

We consider focus groups as a near prerequisite before developing quantitative studies on new ideas and new products. After focus groups or one-on-ones are completed and analyzed, the derived information has its own value and aids in ensuring that national questionnaires are well-formed, using both language and attitudes that guarantee meaningful results.

^ top

A: Telephone, Internet, and/or mail surveys tailored to client's needs and information desired

Since statistical research is expensive and has inherent time constraints, the best research is streamlined, asking only what information is desired and useful in language meaningful to the audience.

Defining specific population sets and surveying methodology are dependent on the information desired and the client. Parks Associates uses telephone, Internet, and mail surveys but prefers either online research or telephone-based surveying to ensure good responses with statistical value. 

Telephone surveys allow for validation (ability to listen in to surveying at the rate of 10% - 15%), which helps us assure that the approach remains constant. In addition, we can pre-test our surveys quickly and then correct any misunderstood questions should they occur. Mail can also be pretested, but it is slower.

^ top

A national sample is a survey that has enough participants so the results are considered representative of the U.S. population as a whole and can be matched to the most recently available census information about US households. Random digit dial is one method helps surveyors meet the requirements for this representation. The traditional sample size to provide adequate coverage of various U.S. family characteristics is between 995 and 1006. The surveyors determine the appropriate breakdown of households to ensure that the results are representative across income, marital status, age, and ethnicity. The sample number suggested provides a +/- 3% statistical validity rate. To get a better rate than that, say +/- 2.5%, requires a substantial increase in survey numbers and is typically deemed too expensive while providing only minimal benefits.

^ top

Using a national sample doesn't provide a survey volume robust enough to determine valid differences in the specific population. Say, for example, that we determine that 20% of US households have a bilingual household member. In a sample of 1,000, we'll then have 200 bilingual households. Breaking down that population into segments creates sets too small for validity. By oversampling, we ensure that analysis further down the line has validity. For example, if 30% of bilingual households have children between 10 - 18, our national sample only gives us 60 households to look at in that category. Validity goes down at that point. By oversampling, we'll have 120 households to measure. This provides us with higher statistical validity.

^ top

There is always a cost/value trade-off. If money were no object, then high-volume surveys would be considered more often. Since this situation is not so, the balance lies between completing enough surveys to provide valid data at a reasonable cost. Four hundred surveys within a smaller segment (millions versus tens of millions) typically provide a +/- validity of 4.5% to 5%. That is good enough. The jump to 3% typically takes that sample back up to 1,000+. The cost per survey is determined by method, difficulty of finding a sample population, and length of survey. Clearly, finding neurosurgeons or bilingual households is a more expensive endeavor than surveying US households in general. That is one reason to use a national sample to determine exact demographics and then pinpoint homes with a greater probability of matching those demographic characteristics.

^ top

Conducting well-done research about existing or past facts is difficult but clearly obtainable. Using the correct methodology with clear questions and solid telephone surveys produces high-quality results. Going beyond stating a "fact," that is, xx% of people bought y-product, requires industry and distribution knowledge for proper analysis.

Where the issue gets even murkier is about purchase intentions and attitudes towards new product and service concepts not yet diffused in the population. Here a knowledge of history and patterns of consumer adoption for product types compared to stated intentions are critical.

The need to marry industry knowledge with research methodology savvy is essential and at the very core of Parks Associates' beliefs. All too often, an industry company understands its industry but completes flawed research because it has no expertise in that arena. On the other hand, using skilled researchers with no industry knowledge for analysis is equally dicey in terms of results.

For example, looking at numbers and facts only in the work-at-home, two-or-more telephone line population, the data alone makes it appear that the percentage of work-at-homers actually went down between 1994 and 1996. Furthermore, the FCC data makes the number of dual-line households smaller than many researchers, including Parks Associates, uncovered. What has caused this disparity? The answer is that between 1994 and 1998, the RBOCs abandoned policing the tariffs charged within the home. Now, nearly all home lines are deemed residential and charged residential tariffs, whether or not the second line is used for business. In the early 90s, the BOCs screened for this so that it could charge higher, more profitable rates for commercial lines where possible. Competition and the desire for incremental revenue changed this. If your researchers do not know their industries, you couldn't get the background for proper analysis.

^ top

Making questions uniformly understandable is time-consuming and requires identifying all possible consumer misconceptions or misinformation. What is high-speed access? How many people know the speed of their modem? In addition, there is always more information desired than time available. Efficiency of wording is critical. For example, asking:

Do you have a PC at home? If yes, then…
Do you have more than one PC at home? If yes, then… and so forth takes more time than:

How many PCs do you have in your home?

These nuances are important in every question. Consumer research is inherently expensive, and it needs to be managed efficiently.

^ top

Check out our CONTACT US page for specific email addresses and our mailing address, call 972.490.1113, or email sales@parksassociates.com

^ top

Subscribe to PARKS POINTS  |  View FAST FACTS
© 1998-2008 Parks Associates. All rights reserved  |  CONTACT PARKS 1-800-PARKS11 or 972-490-1113
    Design: FigDesign  |  Browse Parks Associates Industry Reports