Frequently Asked Questions – MRI Methodology

A: MRI subscribers need look no farther than the paper codebook: the wave dates for a given study can always be found at the top of Page 1 of Section I of the codebook for that study. For those who don’t have a paper codebook handy, the answer can be found at http://www.mediamark.com/mri/docs/wavedates.html.

Q: How many waves of data are included in an MRI study?

A: This depends on the study in question. Fall MRI and Spring MRI are both two-wave studies; Doublebase MRI is a four-wave study. Teenmark and Twelveplus are four-wave studies (as they are based on a recontact to selected households from four waves of fieldwork).

Q: How do I find out which specific waves of data are included in a given MRI study?

A: MRI subscribers will find this information on Page 1 of Section I of the paper codebook. The answer can also be found at http://www.mediamark.com/mri/docs/wavedates.html.

Q: What information is available on the MRI pocketpieces, and where can I get them?

A: There are three MRI pocketpieces for each Spring or Fall study. The Magazine Audience Estimates pocketpiece contains a list of the MRI measured magazines with audiences, median age and income, circulation, and readers-per-copy numbers for each. The Cable Table contains the list of MRI measured cable networks with median age and income, as well as composition by gender, age, prof/managerial, 1+ years of college and presence of children. Cyber Stats contains standard demographic composition (gender, education, age, employment, census regions, HHI, occupation, county size, marital status HH size and presence of children) against Internet users and users of specific online services. MRI distributes printed versions of these pocketpieces to all clients; they can also be downloaded for free from the MRI Plus web site (http://www.mriplusonline.com). Just register for the site, enter, and click the link to “pocketpiece data.”

Q: How can I find out the release date for the next MRI study?

A: Because the official release date for a given MRI study depends on fieldwork completion milestones (which are notoriously difficult to predict), as a matter of policy, MRI does not announce official study release dates more than a couple of weeks in advance. MRI maintains a list of official release dates, as well as a list of last year’s official release dates at http://www.mediamark.com/memri/Release_Schedule_2003.htm. If you have additional questions, call the MRI Client Services Hotline at (212) 884-9299 or your MRI Sales Representative.

Q: What is the release sequence for an MRI study?

A: In the case of a two-wave (Spring or Fall) study, MRI releases the demographic and media data first, as this information is available immediately following completion of the personal interviews. A few months later, after MRI has stopped accepting questionnaires and finished processing the results, the bulk of the questionnaire data, mostly product data, is released. Generally, the product data is followed within a few weeks by data that requires special weighting, such as Buying Styles, VALS and Technographics. Finally, a few weeks after that, Consumer Innovators and Leisure Styles are released, completing the release. The entire sequence usually takes about seven months.

Q: What studies compose a Doublebase study?

Doublebase is always released following the Spring study, and comprises the waves from that Spring study combined with the previous Spring study. For more information, consult the bottom of the Wave Dates page, http://www.mediamark.com/mri/docs/wavedates.html. There you’ll find the specific waves that went into each Doublebase study.

Q: How do I know which MRI study to use?

A: The major determinants of which study to use are the status of the most recent studies (that is, where in the release cycle they are), the type of data one needs, and the importance (or lack thereof) of a large sample size. If a large sample size is required (because, for instance, you’re using a narrow target or examining an activity with a low incidence) or brand volume data is part of the equation, you should use a Doublebase study. If the target or activity is large enough to remove sample size as a consideration, you should consider the most recent Spring or Fall studies, as those are drawn from more recent fieldwork. In short, the best study will be the most recent one that has all the data you need.

Q: How do I find the actual wording of a given MRI question?

A: For a full index of the most recent available product questionnaire, go to http://www.mediamark.com/memri/Documentation.htm and click the link to “Browse the Current Product Questionnaire” under “MRI Data – Methodology” to find an index to allow you to either drill down to a given page or search for an item within the latest available product questionnaire. The pages themselves are published in Adobe Acrobat form. Naturally, this requires that the Adobe Acrobat Reader (which can be found at http://www.adobe.com) has been installed on your computer.

Q: What does MRI ask in the personal interview?

A: The MRI personal interview covers most of the media measurement (the exception being specific TV programs) in the MRI survey. Prominent among the media measured in the personal interview are magazines, national newspapers, radio, TV (overall usage of TV and TV daypart usage, as well as cable networks, but not individual programs) and the Internet. The MRI interview is also where MRI obtains all demographic information. The third major element is business-to-business decision making, but other lesser-known items include personal computer and phone ownership, vacation time-share ownership and usage of van lines for moving.

Q: What does MRI ask in the product questionnaire?

A: The majority of the MRI product book is devoted to usage of products and services, but it also contains TV program viewership, psychographic and lifestyle questions. MRI provides printed copies of each wave’s questionnaire to all MRI clients. A full index, with links to the pages in Adobe Acrobat format, is also available on the MRI web site. Just go to http://www.mediamark.com/memri/Documentation.htm and click the link to “Browse the Current Product Questionnaire” under “MRI Data – Methodology.”

Q: What is MRI’s response rate?

A: Response rate numbers vary each wave. Generally, MRI gets interviews in about 65-70% of the households approached. Of the interviewed households, about 60% return product books that are complete enough to be accepted.

Q: What is ascription?

A: If an interviewed respondent does not return a completed product questionnaire, MRI ascribes his or her product usage, TV program viewership and lifestyle (but not psychographic) information. This means that when MRI is confronted with a respondent who does not complete the product questionnaire, we find the closest demographic and geographic match who has completed the questionnaire and copy that person’s questionnaire data to the record of the respondent who has not completed the questionnaire. While this may seem absurd on an individual level, it is highly accurate on the aggregate level. Ascription using demographics is effective for accurately predicting product usage and TV program viewership; it is not accurate for psychographic data sets, such as Buying Styles, VALS and Technographics – for these, MRI creates custom weights.

Q: What is the confidence level for an MRI study?

A: Technically speaking, MRI will not publish a magazine audience number unless it can guarantee a 95% confidence level based on a relative 2-sigma value which is less than 25% of the audience measurement for that magazine. In other words, MRI guarantees that any reported magazine audience number, if re-measured 100 times, would in 95 of those 100 cases fall within 12.5% of the number given.

Q: How does sample-balancing work?

A: No matter how well designed, any sample of a population is likely to differ somewhat from the population being measured, and MRI’s sample of the adult population of the 48 contiguous U.S. states is no exception. In order to correct for differences, MRI uses a statistical method called sample-balancing, which is designed to remove demographic differences between the MRI sample and the projected population by adjusting the relative weights of each individual according to his or her demographic characteristics. Benchmarks for these weights are provided by the U.S. Census Bureau. For MRI’s official description of its sampling and sample-balancing procedures, consult http://www.mediamark.com/mri/docs/natnlsurvey.html.

Q: How does MRI measure television?

A: MRI provides three separate measures of television: (1) total usage of television by daypart, (2) weekly hours viewed of particular cable TV networks, and (3) viewership of specific TV programs. To measure total usage of television, MRI asks respondents how many half-hours of TV they watched (regardless of whether it was network, cable, syndicated or local TV) during defined time periods on the most recent Saturday and Sunday. For cable networks, MRI shows a logo card for each network, and, if the respondent indicates that he has viewed the network in the past 30 days, he is asked how many hours he or she has watched in the past 7 days. The TV program information comes from the product questionnaire: respondents indicate whether or not they viewed the most recent airing of each program, and, if they have, they are asked where they watched it and how much attention they were paying.

Q: How does MRI measure radio formats and radio dayparts?

A: MRI does not ask respondents about radio formats or networks. MRI asks respondents which stations they listened to during defined dayparts, and, using industry information, MRI can use that information to determine to which formats and networks the respondents were exposed.

Q: How does MRI measure magazine readership?

A: During the personal interview, MRI uses a four-step process to measure magazine readership. The first step, called a six-month screen, eliminates those publications that the respondent has not read in the last six months. In the second step, the respondent is then asked about the frequency of reading (on average, out of four issues published, how many does he or she read?). The third step determines whether the respondent qualifies for the publication’s average issue audience, determined by whether or not he or she has read the publication during its most recent publication period (for example, if the magazine is published weekly, has the respondent read or looked into it in the past 7 days?). In the final step, the respondent is asked about more detail (where read, how obtained, etc.) on the magazines for which he or she has qualified as part of the average issue audience. For more on how MRI measures different media, consult the MRI Essentials Quick Sheet at http://www.mediamark.com/MEMRI/quicksheets/Essentials_Quick_Sheet.pdf.

Q: How does MRI measure daily newspaper readership?

A: Starting with a pre-list of national daily newspapers, the interviewer inquires whether or not the respondent has read them in the last 7 days; the respondent is also invited to tell the interviewer any other newspapers he or she has read during the same period. For any daily newspapers read in the last 7 days, the respondent is then asked about frequency of readership (on average, out of five issues published, how many issues does he or she read) and whether or not the publication was read yesterday (which determines average issue audience). If the respondent qualifies for average issue audience, he or she is asked where he or she read it, and how it was obtained. For more on how MRI measures different media, consult the MRI Essentials Quick Sheet at http://www.mediamark.com/MEMRI/quicksheets/Essentials_Quick_Sheet.pdf.

Q: How does MRI measure Sunday and weekend newspaper readership?

A: Starting with a pre-list of national newspapers with Sunday and weekend editions, the interviewer inquires whether or not the respondent has read them in the last 4 weeks; the respondent is also invited to tell the interviewer any other Sunday or weekend newspapers he or she has read during the same period. For any Sunday or weekend newspapers read in the last 4 weeks, the respondent is then asked about frequency of readership (on average, out of four issues published, how many issues does he or she read) and whether or not the publication was read the most recent Saturday or Sunday (which determines average issue audience). If the respondent qualifies for average issue audience, he or she is asked where he or she read it, and how it was obtained. For more on how MRI measures different media, consult the MRI Essentials Quick Sheet at http://www.mediamark.com/MEMRI/quicksheets/Essentials_Quick_Sheet.pdf.

Q: How does MRI measure Sunday supplements (Parade, USA Weekend, etc.)?

A: MRI does not ask respondents about specific Sunday supplements. If the respondent qualifies for the average issue audience of a Sunday or weekend newspaper known to contain a given Sunday supplement, then the respondent is included in the average issue audience of that Sunday supplement.

Q: How does MRI measure outdoor?

A: MRI’s measure for outdoor is based on the respondent’s estimate of the number of miles he or she has traveled as a driver or passenger in an automobile.

Q: How does MRI conduct the Teenmark fieldwork?

A: During the personal interview for the adult (18+) study, MRI asks the name and age of all individuals in the household. After the conclusion of the adult fieldwork, MRI mails a questionnaire directly to all teenagers identified in the two waves which make up the Spring study. Included in the questionnaire are all the media, product usage and psychographic questions that MRI reports in the Teenmark study; there is no personal interview for the teens. MRI reports the data points in common between the past two questionnaires (this year’s and last years), producing a four-wave teen recontact study.

Q: What’s the difference between Teenmark and Twelveplus?

A: Teenmark contains all the questions we ask the teenagers; Twelveplus includes only those things that we ask both the teens and the adult respondents. Because the Twelveplus study includes both teenage and adult respondents, it has the largest sample of all MRI studies, bigger than Doublebase, which only includes adults. However, because the questions must be asked in both the teen questionnaire and the adult survey, it has the least data.

Q: How is the MRI Twelveplus study compiled?

A: By combining all the compatible data gathered from the teen and adult fieldwork. The data on 18-year olds and 19 year-olds gathered from the Teenmark questionnaire is excluded in favor of the higher-quality data gathered from the 18 and 19 year olds in the adult survey.

Q: What is the definition of “head of household”?

A: The respondent is asked to designate which person in the household is considered the “head of household” during the personal interview. No explanation or definition of “head of household” is given; the respondent is expected to interpret the phrase and apply it as he or she sees fit.

Q: What is the definition of “homemaker”?

A: In the MRI Product Questionnaire, the “homemaker” or “principal shopper” is defined as “the person who does most of the shopping for groceries and other household items.” That terminology determines who is given the task of answering the “homemaker” section of the questionnaire, which includes “products used by or for the entire household.” Such products include (but are not limited to) food, paper products, baby products, cleaning products, and pet products.

Q: What is a “respondent homemaker,” and why is this significant?

A: As the name suggests, a “respondent homemaker” is someone who not only participates in the personal interview, but also fills out the homemaker section of the codebook. When the homemaker is also the respondent, all the data in the adult survey is drawn from the same person. Marketers consider homemakers (also called “principal shoppers”) to be the point-of-purchase decision-makers for items that are purchased for the entire household. To properly compare the purchase decisions of these people with their media habits, it is common practice to restrict one’s base to “total homemakers,” excluding both the media data contributed by non-homemakers and the household purchasing data contributed by homemakers who did not participate in the personal interview.

Q: How do I code for all college graduates?

A: The education questions are asked by “highest degree received.” Therefore, you can isolate college graduates by “netting” or “orring” together the codes for Bachelor’s degree (an Associate degree is not considered enough to qualify) with all the post-graduate degrees (Master’s, Professional degree and Doctorate). In other words, “all college graduates” includes those whose “highest degree received” is either a Bachelor’s degree, a Master’s degree, a Professional degree or a Doctorate.

Q: When I combine those with “full-time” employment status and those with “part-time” employment status, I only get about 65% of the population. Why is that number so low?

A: Employment statistics only include those who are either actively employed or seeking work; in other words, students, retirees, full-time homemakers and those who are not seeking work for reasons of disability are excluded. These groups make up about 35% of the adult population, leaving only 65% of the adult population to be included in the job market.

Q: How do I code specifically for top management?

A: In the demographics section (Section III) of any MRI national study (Spring, Fall or Doublebase) codebook, under “Respondent Job Title” MRI provides several codes for isolating top managers: President, Vice President, Treasurer, Chairman of the Board, Member of the Board, Comptroller, General Manager, Owner-Partner, Manager and “Other (foreman, supervisor, administrator, superintendent).” There is a very useful summary code for “Top Management” defined as “Professional/Managerial with IEI $35,000+ and job title code 1 through 8,” referring to the titles listed here from President through Owner-Partner.

Q: How are A, B, C and D counties defined?

A: A, B, C and D counties are defined and designated by Nielsen. We print the Nielsen definition in the demographic section (Section III) of our national study (Spring, Fall, Doublebase) codebooks. As of this writing, “A counties” were defined by Nielsen as “all counties in the 25 largest metropolitan areas.” “B counties” were “counties with over 150,000 population that are not in Class A (25 largest metropolitan areas) plus counties that are part of the metropolitan area of cities in such B counties.” “C counties” were “counties not included under A or B having over 35,000 population plus counties that are a part of the metropolitan area of cities in such C counties.” “D counties” are everything else.

Q: Under “race,” I only see “white,” “black,” “Asian,” and “other.” Why isn’t “hispanic” an option?

A: The U.S. Census Bureau does not treat “Hispanic” as a race, and neither does MRI – race and “Hispanic” are two independent characteristics. Race is “judged” by the interviewer at the end of the personal interview. Qualification for “Hispanic” in the MRI fieldwork is determined by the respondent’s answer to the following question: “Do you consider yourself to be of Spanish or Hispanic origin or descent?” In every wave, there are people who qualify as both “black” and “hispanic”, there are also cases of people who qualify as both “Asian” and “Hispanic.”

Q: How do I find out which areas are included in a specific Census region, subregion or marketing region?

A: This information can be found in Section III of any MRI national study (Spring, Fall or Doublebase) codebook, in the section entitled “Definitions of geographic areas.”

Q: Is it possible to isolate specific media markets using the MRI national study?

A: Yes, it is. Since MRI intentionally oversamples the top ten Nielsen DMAs from the time when our sample was designed (early 80’s), we are able to publish these in a special “Local Market” add-on to the Doublebase study. The MRI Local Markets include Boston, Chicago, Cleveland, Detroit, Los Angeles, New York, Philadelphia, Dallas/Ft. Worth, San Francisco and Washington, D.C.

Q: What does MRI Local Market data include?

A: MRI Local Market codes allow subscribers to run any MRI national study data point against the local market base. Also included are local media (such as local newspapers, TV and radio stations) that are not included in the national study, as well as local retailers.

Q: The MRI local markets do not match the present list of top ten Nielsen DMAs. Why not?

A: The top ten local markets were built in to the MRI sample frame when the MRI sample was originally designed, ca. 1980; Nielsen has since revised their list many times. To change the local markets in MRI would be to redesign the sample, an expensive process that is not warranted at present. So we’ve kept the list constant since the beginning.

Q: What is an “MRI standard category”? What is a “non-standard” category?

A: The concept of an “MRI Standard Category” was born out of the recognition that certain products (particularly traditional packaged goods) follow a consistent pattern of usage. Toothpaste and Orange Juice, for instance, share certain qualities: those people who use them can be expected to do so constantly, and to be able to estimate the number of times they do so during a week. In both categories, there are different types; in both categories, there are a variety of competing brands. Non-standard categories include everything that doesn’t fit such a mold, such as big-ticket items (cars, personal computers) and services (financial services, insurance). Those categories considered “MRI Standard Categories” are not a majority of the categories MRI measures; they are only the single biggest type.

Q: In a standard category, how does MRI determine the usage breaks between heavy, medium and light users?

A.: These breaks are determined by creating a distribution of recent usage at the time that the category is initially included in the MRI national study. MRI divides users into three distinct groups that are as even as possible, with those users who indicated the highest usage levels combined into the “heavy” group, those with the lightest usage combined into the “light” group, and those in between combined into the “medium” group.