Question

CASE 2.5 Cosmopolitan Magazine 373 CASE 2.5 cOSMOPOLITAN MAGAZINE Surveys are often sponsored and conducted by various popular press outlets to spur interest among and create a more inreractive relationship with their readers. Nowadays, surveys asking people abour their consumer preferences, shopping behavior, general opi nions, even their sexual habits, are commonplace, but this was not always so. A land mark survey of this kind, often characterized as such due to its ens sample of 106,000, was conducted in 1980, by Cosmopolitan magazine. The tremendous response to this survey and the extended publicity that it generated caused other publications to take note; soon after, all manner of newspapers and magazines were including surveys within their pages, each geared to their particular readers' interests. Cosmopolitan, a monthly lifestyle magazine that is targeted to a female audience, is published in more than 25 languages and sold in more than 40 countries. Cosmopolitan typically covers topics such as: Relationships, careers, personal growth (e.g., personality quizzes), and female scxuality Fashion, beauty care secrets, and tips on fitness and staying healthy Profiles of celebrities and broad coverage of pop culture Despite the female target audience, the magazine has a large niche of steady male readers (14.7 percent). The median age of Cosmopolitan readers is 31.4 years (31.8 per cent in the 18 to 24 age group, 26.3 percent in 25 to 34, and 27.9 percent in 35 to 49). Median individual income is just more than $26,000; median household income is about $58,000, with 66.2 percent of the readership earning more than $40,000 per household. Sixty-five percent of the Cosmopolitan audience has some college education or a college or higher degree. Seventy-two percent are employed outside the home, with 54 percent employed full-time. Forty-two percent of the readers are unmarried, 40 percent married, and 19 percent divorced, widowed, or separated. The magazine also claims that its readership is more involved than that of similar publications, spending an average of 80 minutes per issue compared with 50 to 73 minutes for readers of competitors. The decision to include a survey on sexual behaviors of American wo the Cosmopolitan issues was made in the summer of 1979. A total of 79 multiple choice questions were created and compiled by a panel of Cosmopolitan editors, grouped into three sections covering personal background, sexual experience, and sexual manners and morals. The survey was run as a regular Cosmo article in the January 1980 issue rather than as a perforated tear-out with return postage guaranteed. In what was an unusual request for this sort of undertaking, the reader was asked to completea fairly long questionnaire and send it at her own expense. The questionnaire was accom- panied by a letter from the editor which stated the purpose of the survey and exhorted the reader to "share her sex life" with other Cosmo readers. men in one of Simmons Market Research Bureau, a public-opinion research group located in Manhattan and which served many other magazines and publishers, was commissioned to coordinate the data collection and analysis. The task of interpreting the analysis c piled by Simmons was assigned to Linda Wolfe, behavior and science writer and author Linda Wolfe was also responsible for enriching the purely statistical findings from the Simmons analysis with real-life testimonies, by reading the many unsolicited messages and stories that were sent in along with the questionnaires. Linda Wolfe stated that she was deeply interested in the sexual changes that had taken place in America since her college days in the 1950s, but she had given up hope of ever being able to write about them until she came across the Cosmopolitan sex survey When she heard that 106,000 women had responded to the survey, her interest picked up immediately. She compared this enormous sample size with those of previously conducted studies. Hite, in the Hite Report, had studied 3,000 women. Gay Talese had spoken with at best a few hundred. Even Alfred Kinsey had relied on a sample of only 5,940 partici- pants for the famous "Kinsey Report," Sexual Behavior in the Human Female ever, Wolfe said that initially she was reluctant to accept the assignment of interpreting the 106,000-response survey for Cosmo because she believed that, although portant, respondent variety and representativeness were also critica! e survey's findings. She did not want to write about the sex-related haviors of just the "Cosmo girl," a typical reader of the magazine, something scientifically valid and up to accepted standards of rigor. After an ons executives, Wolfe was struck by how varied a sample of survey had tapped. She was informed by Val Appel, research mons, that the (women) respondents were between the ages of 14 and but to do appointment with Simm im every geographical region of the country, and ranged in occupa ident and college professor to motel proprietor and telephone us, although all respondents had in common the fact that they read the ssue in which the questionnaire appeared, their geodemographic differ ences conferred to the sample remarkable diversity. This diversity was the foundation of the argument that the sample represented American females and could be used to under- stand how the sexual revolution affected the values and behavior of American women at Co It was decided that the material gathered was adequate to produce an article, called "The Sexual Profile of That Cosmopolitan Girl" and published in the September 1980 issue, as well as a book, titled The Cosmo Report, published soon after in 1981. The two publications concerned two different groups of women. The article analyzed the sexual habits and attitudes of only a portion of the survey's respondents, some 15,000 women who most resembled the magazine's typical reader: a woman between the ages of 18 and 34, who lives in a city of more than one mllon and earns her own living On the other hand, the book examined all 106,000 respondents, a sample comprising a much wider sector of American women. This sample composition is described as follow ing in The Cosmo Report book, " while 85% of the respondents are between the ages 18 and 34, some are as young as 14, and some are in their forties, fifties, sixties, and even seventies. Although a fifth of The Cosmo Report women come from cities of mor than 1 million, another two-fifths come from smaller cities, and the remainder from sul urban and rural areas. And although the women in The Cosmo Report are primaril women who earn their own livings, chiefly as managers, administrators, professional technicians and office workers, there are also some homemakers and numerou students." Later the author commented: "Having worked closely with the Cosmo data, I myself suspect that the sexual practices of this largely 18 to 34 year old sample is not unusually extreme, and that the Cosmo women may, in fact, be quite representative ot young American women as a whole." The publication of the survey results generated great publicity for the magazine Cosmopolitan's editor, Helen Gurley Brown, explained: "After all, this was the biggest response to any magazine survey in history and surely the largest sex conducted." Television programs such as the Today show, the Tonight Show, and Phil Donahue, among Linda Wolfe, to investigate and discuss the results of the Cosmo sex sur survey ever many others, all hosted Cosmopolitan editors as well as the writer, vey Questions 1 What is the population definition of the survey? What sampling frame was used? Were 2 Bearing in mind Linda Wolfe's statements about the sample size and the diversity of these appropriate choices? If not, what would be? the respondents, do you believe the sample was representative of the Cosmopolitan audience? Of the American female population? Explain your answers 3 Are there key differences between diversity and representativeness? How might the magazine have gone about (a) collecting a more representative sample of each of these groups, or (b) statistically checking and analyzing the existing data to render them more nearly representative? What statistical methods might they call on for such purposes? 4 What sampling procedure was used? List and discuss in detail its advantages and disadvantages. Would the resulting sample be adequate to extrapolate to the readership of other, similar magazines? To the population of Cosmo reeders who chose not to respond? If you had to guess, in which directions were certain key measures biased if they were at all? 5 Consider the description of the questionnaire design and data analysis interpretation, and identify potential problems. What would you recommend to the magazine to mitigate such problems in the future? Would any of them likely have been less severe if the magazine had remunerated respondents for their participation, or at least for additional, standardized demographic data? 6 Suppose that you were hired by the magazine to design this survey. Develop a detailed survey proposal; among other elements, include information on sample design, sample generating procedure, and type of the survey that you would propose to the magazine. Suppose that the magazine's information needs might have a longitudinal component; how might one handle that while ensuring respondent anonymity?

Answer 1

Cosmopolitan

Answer 1: When a scientific examination is conducted for a given population by compiling data or information from a sample that is chosen from the given population and is presumed to have similar characteristics as that of the population, thereafter analyzing the data using statistical tools and techniques, is the population definition of survey.

Sampling frame used was the female readers of Cosmopolitan magazine.

It was found that the sample distribution of female readers of the magazine was diverse and varied hence it was considered to be the best possible representation for the American women for the survey.

It would have been better if a statistically proven sampling technique was implemented for better representation of the population.

Answer 2: Sample was not the representative of the Cosmopolitan audience as the magazine also has male readers. But it could be considered as a representative sample of the American female population because of its size (106000) and geo-demographic diversity. There might have been females who do not read Cosmopolitan and have not been considered in the survey, but assuming that they must have some similar geo-demographic characteristics with that of the huge sample size, the sample is assumed to be representative sample of American women. It would have been more authentic if sampling was done based on these geo-demographic differences.

Answer 3: Diversity and representation are two different terms that describe a sample. The representativeness of a sample is based on the objective of the survey and the population considered. Diversity is an essential component for a sample to be representative. The magazine could have opted for more scientific approach like stratified sampling technique. The population could have been divided on the basis of age, education or profession. Further samples could have been derived from those subsets.

Answer 4: It was evident from the case that in this survey Convenience Sampling Method was used.

Advantages of Convenience Sampling Method

1. Convenient and hassle free.
2. No extra cost or time involved.

Disadvantages of Convenience Sampling Method

1. Very low chances of true representation.

No the resulting sample will not be adequate for extrapolating to the readership of other similar magazines because of varying population and hence sample frame characteristics. It can be extrapolated to the female readers of the magazine who did not respond. They key measures were biased only towards the magazine readers.

Answer 5: The questionnaire was long that consisted of 79 multiple choices questions. The respondents might lose their interest in answering the questionnaire till the end. The questions were created and compiled by the Cosmopolitan editors and not by any agency or person specialized in conducting research. The structure of the questionnaire is not clear as personal information and demographic information was put under one heading “personal background”. The way of collecting back the responses at the expense of the reader was not justified. The task of interpretation was given to someone who had expertise in behavioral writing and authoring.

To mitigate such problems in future they can outsource expertise in questionnaire designing, checking the validity and reliability of the questionnaire by scientific method followed by appropriate methods of collecting response. The interpretation should be done by a specialized researcher who could interpret the statistical results obtained after the data analysis.

Remunerating the respondents would not have reduced the severity of any of the above mentioned problem because the major loophole was in the data collecting instrument and data interpretation.

Answer 6:

Survey Proposal Design( Broad Outline)

Introduction

1. Back ground of the survey
2. Identification of the objectives
3. Expected data
4. Identification of population
5. Expected sample size
6. Identifying the data collection instrument and data collection method

Body of the proposal

1. The duration of the survey specifying the start and the end date.
2. Longitudinal survey can be avoided to ensure respondent anonymity.
3. Questionnaire design- Prepare the outline, how many questions, type of questions, pilot study to check the validity. Reliability testing. (Outsourced)
4. Specifying the sampling technique- Stratified sampling technique.
5. How to collect the responses.
6. Availing data analysis and data interpretation services.(Outsourced)
7. Explaining the usage of the results and data obtained.