il‘ III I III (III I I I I I III THS A COMPARATIVE STUDY OF THE EFFECTS OF LIVING HABITS UPON RADIO LISTENING HABITS Thesis for fhe Degree of M. A. MICHIGAN STATE COLLEGE Eager B. IIIamIin $61.50 ‘II‘IQIIIIIII'III‘ This is to certify that the thesis entitled presente by wake/64 has been accepted towards fulfillment of the requirements for WW de rec in W 9 28¢“ W's. Major professor OVERDUE FINES: 25¢ per day per item RETURN!“ LIBRARY MTERIALS: Place in book return to remove charge from circulation records TIT A COMPARATIVE STUDY OF THE EFFECTS OF LIVING HABITS U'PON RADIO LISTENING HABITS by ROGER B. HAMLIN A THESIS Submdtted to the School of Graduate Studiee of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER or ARTS Department of Speech, Dramatice, and Radio Education 1950 Chapter 1. Chapter 2. Chapter 3. Chapter 4. Chapter 5. CONTENTS ‘?*:1r;:‘a ‘vivl‘ ' . ,I“ 4 v.‘:' INTRODUCTION METHODS STATISTICAL ANALYSIS CONCLUSIONS APPENDIX ACKNOWLEDGMENTS Grateful acknowledgment is made to Dr. Melvin Springer. Mathematics Department, Michigan State College. flor his infin- ite patience and generaus assistance in working out the sta- tistical analyses of the findings of this thesis. The project was followed with careful concern by Dr. Springer. and it was only after thorough testing of the findings that results were established as significant or chance relationships. Mr. William Kohler, Editor of Sponsor magazine, New York City, origanally suggested the possibility of a survey of the effect of living habits upon radio listening habits. The writer is indebted to Mr. Kohler tor the original suggestions that prompted this thesis. Mr. Joseph Calaway, Director of Radio Education, Michigan State College, has been a valuable guide and source of inform- ation in assembling the material, and it is to him that much of the credit for the 'faith.to persist“ is due. Dr. Donald Hayworth and Mr. James Tintera of the Depart- ment of Radio, Speech, and Dramatics of Michigan State College have aided in reading the thesis and offering helpful sugges- tions and corrections. 1 “III Chapter 1. W Chapter 1. INTRODUCTION The measurement of public preference for radio programs has long been an accomplished science. Thanks to Hooper rating and other kindred analysts of public opinion, it is now possible to determine the stage of popularity or de - clrne of any program on the air. Accordingly, such waste of money and public interest has been avoided. It has also been a great boon to advertisers supporting such programs to know when their'broadcasts were being accepted and to what extent such programs compare with others in popular acceptance. However. little concern has been given to the question of'what kinds of programs appeal to different types of indi- viduals. People do not listen abstractedly merely for the sake of listening: people of all walks and climes are apt to beflgelective. Their selections, it might reasonably be as- sumed, may be determined by their'backgrounds and their char- acteristic manner of life, frequently referred to as "their mental set."l Hypothetically, people who have little education might prefer programs that require little thought or reflec- tion. Assumedly. people of higher income brackets might be 1 Matthew N. Chappell, a C. E. Hooper, Radio Audience Measurement, Stephan Days. New York, 1944, p. 26. inclined to favor programs of a more literate or classical level, such as classical music, discussions and forums. and drama of the more thoughtful nature. It might be conjectured that people of a low scale-economic level would favor programs such as the ”soap opera". “give-away”. or raucous comedy. But these are merely assumptions drawn from observable data found in other social phenomena in the world around us. Accordingly, the purpose of this research shall be to conduct a study of a representative metropolitan city to de- termine the\measure and the manner in which living habits affect listening habits. Are people predisposed to listen to certain types of programs because of their manner of living. or is listening purely an individual matter subject to human personality variations, as varied and complex and unpredicta- ble as human genes? If individual differences and personal conditioning are the sole determinants of individual listen- ing, then Hooper rating of programs would undoubtedly convey the obvious answer as to what should be presented on the air. But if people listen to different radio programs because of their living habits, it might then prove profitable to offer programs on the basis of the social complexion of an area or community. In this study we shall attempt to answer this problem by posing the following questions: I I. In regard to such programs as classical or semi- classical music, hillbilly music, dance music, mystery dramas, serious dramas, serial dramas (soap Operas), comedy, variety shows, shopping programs, discussions and forums, give-away programs, and religious devotional programs, how do the following react: A. B. C. D. E. Church-Coors. People of varying degrees of education, such as college graduates, high school graduates, gram- mar school graduates only. People of different occupational levels, such as the professional, the white-collar, and the laboring groups. People of different-size families, such as the family of two adults without children, the family with one child, and the family with two or more children. People of varying incomes, such as incomes of $5000 or above, of 33000 to $5000, of 82000 to $3000, and those of 82000 or below. 11. What may be learned about the individual tuning habits from the Radio Diaries left for a week in each home that was interviewed? Such a study shall seek to investigate the following: A. B. C. D. Is there a tendency toward greater selectivity in choosing programs, varying between the college graduate, the student of high school level, and the grammar school graduate? Is dial-hopping more prevalent in one group than in another (i.e. in the college, high school or grammar school group)? Is there a tendency to follow a regular listening pattern throughout the week? In which group is this more prevalent: college, high school, or grammar school level? What is the average number of listening hours for all classes and groups? In which is this the greatest or least, the higher educational levels as compared with the lower? III. For the purpose of commercial interest in advertising, an attempt was made to determine the amount of public ac- ceptance or rejection of radio advertising. A study of similar’nature, conducted by Clifford Kirkpatrick in 1933, in the city of Milwaukee, revealed a great public distaste and even hostility toward radio advertising, which the writer assumed might be more or less expressive of public 2 reaction to radio advertising at large. The question- airro employed in the present research has attempted to investigate the matter of public reaction to radio adver- tising. Accordingly, it has been asked which group gives greater acceptance or rejection to radio advertising, the -higher or lower educational level, or the higher or lower income level. How many radio listeners have been influ- enced in their buying by radio advertising? In which group is the influence greater? I It is hoped that such a survey may reveal the extent to which listening is influenced by the personal living habits of a people. Can a living pattern for a city or community be correlated with a listening pattern? Would certain living habits naturally predispose toward defi- nits programs? Is it possible to build a suitable program ‘based upon the sociaILcomposition of an area? These are the questions this thesis will attempt to answer. '5 Clifford Kirkpatrick, "Report of a Research into the Attitudes and Habits of Radio Listeners,” stb Book Pub- lising Company, St. Paul, Minnesota, 1933. p. 43._ Chapter 2. was Chapter 2 METHODS Selection of the Sample to be Measured In conducting a study of this kind, it was thought ad- Ibisable to choose a population for study that would yield the typical levels of society representative of all classes and conditions at large, and which would at the same time offer a population large enough for accurate study. Lansing, Michigan, appeared to have the desired charac- teristics for such a study. A city of thirty thousand home units, there is to be found in this metropolis people of all vocations and educational levels in sufficient numbers to make random sampling possible. Lansing is an industrial city, the home of Oldsmobile, Roe, and Motor Wheel manufacturing companies, as well as oth- er small factories. It would thus afford large numbers of homes that would be typical of industrial workers' opinions.3 At the same time it is a college city, with Michigan State College on its eastern border. A good share of the faculty and student body reside in the Lansing city area.4 3 36,470 industrial workers. (Lansing Labor Market.) 4 3.725 faculty members and students. (Estimate, Michigan State College Student Housing Office. No official figure.) Being the capital city of the state, Lansing would offer an interesting cross-section of state employees, consisting of 5 office workers, politicians and specialists. Determinationggf the Kind of Sample to Use The relation between sample size and statistical relia- bility presented a problem. How large a sample would be rep- resentative enough to yield accurate data? Reliable opinions secured advised that a smaple of five hundred would be suffi- cient.6 ‘ Further, the problem developed as to what kind of sample should be used. The choice lay between the pure random sample or a stratified random sample. Obviously, the pure random sample would be too large and impractical for such a survey. Hence, it was determined to use the stratified random sample. A.map of the city was procured from the Chamber of Com- merce, revealing every street, road, and alley, and clearly defining the city limits. It wasdeemed advisable to limit the study to the primary metropolitan area, which would ex- clude the city of East lensing, the town of Pottersville, and the newly-populated environs of the city beyond the city lim- its. The map of the city was then divided proportionately into thirteen districts, an arbitrary zoning for convenience, 65,200 state employees in Metropolitan Area. Mr. William Kohler, Editor of Sponsor Magazine; Mr. Joseph Galloway, Director of Radio Education, Michigan State College. since little data could be secured as to political wards or other city divisioning. It was then determined, after strat- ifying the population into thirteen districts, to take an approximate two percent sample from each of the thirteen dis- tricts, an aggregate sample of 500 at the most. The City Directory was used for reference, revealing every street in the city, every house, its occupant, and his occupation. By use of such reference, homes were chosen at random from each of the thirteen districts. f Securing the 1mm The method chosed for securing the desired information was the use of the personal interviewer, who employed a ques- tionnaire and left a Radio Diary in each home for a period of one week. It wa ught that the personal interview would offer a more accurate approach to the problem of securing personal opinions. Telephone calls are often hasty and impersonal, affording the person interviewed little time for recall or reflection. Questionnaires mailed to homes or individuals frequently fail to give accurate results, in that the matter of personal responsibility or interest is often lacking when one is not answerable to any particular party, or concerned with the results, or compensated for the trouble of filling out the answers to the questions. Accordingly, it was planned that each of the 500 homes would be personally interviewed. A form letter was sent to each home three days in advance of the interview, informing the family that the interviewer would call at the resident's convenience on a specific day. The letter explained that the information sought would offer helpful information to the radio industry at large; also, that the project was being conducted with the endorsement and approval of the Department of Radio, Speech, and Dramatics of Michigan State College, and was signed by the Director'of Radio Education. These letters, sent to the various homes whose names had ‘been chosen at random from the city directory, provided a cor- dial ingress to the home. The interviewer was politely re- ceived, and time generously given to answer the questions on the questionnaire. - The interviewer had a double motive in his visit: He had to secure the desired information necessary for the question- naire; and he had to establish.enough confidence and interest on the part of the resident to persuade him to cooperate in keeping a Radio Diary for one week, in which the person would accurately record the hour and time of radio listening for his entire household. At the end of the week, the interviewer returned to collect the Diary and confirm the answers given on the questionnaire. Almost without exception, answers were thoughtfully and cheerfully given. 10 The questions directed by the interviewer were beamed at the family at large, rather than at individual members. In- quiry was usually prefaced with the explanation that it was desirable to know what the usual practice of the household was. Though individual members of a family might differ in their choice of programs, it was assumed that there is a general preference pattern in the home. The home that eschews mystery and ”long-hair” music consistently, tunes it out. Notwith- standing the fact that some protesting member of the family may like it, and from time to time listen to it, in the realm of averages, it was assumed that that home does not custom- arily tune to mystery stories or symphonic music. likewise, it has‘been assumed that the highest educa- tional level attained in the home will profoundly influence the members of that home. Educational levels may vary in the family composition, but where it is found that one or two members have had college training, or high school education, or other training, there may be found a corresponding in- fluence in program choice. Thus, it was thought advisable to conduct the interviews entirely on the basis of the opinion and practice of family units, rather than on individual pre- ferences. 11 Individual interviewing was conducted for a period of over a year, from June, 1948 to September, 1949. Each interview averaged from fifteen to twenty minutes spent in the home, with an additional five to ten minutes upon the return of the inter- viewer to collect and confirm the Diary. As the project was such a protracted one, involving an extended period, careful concern was given to the disposing of a representative number of the'Idaries throughout the summer listening season as well as the winter listening season. Chapter 3. STATI STI CAL AN ALYSI S 12 Chapter 3. STATISTICAL ANALYSIS elem FICANCE The term significant with its various applications is a statistical term that we shall be using extensively through- out this thesis. When any result is referred to as signifi- ‘gggg,or:not significant, it is implied that the pertinent data has been subjected to a statistical analysis which has revealed the probability that the result occurred from chance alone; only when this probability is less than 0.01 will the result be termed significant. A.word of caution, however, may be in order at this point. One cannot conclude that a percentage difference is significant simply because it is larger than another percentage difference which is known to be significant, for the size of the respect- ive samples that are being compared may give a deceptive ap- pearance to the percentage difference. Fer example, in the educational classifications, it was found that the difference in percentages of high school and college people listening to dance music was 9.5%: this difference was found to be' 13 significant. Again, the difference in percentages of college and grammar school people listening to the soap operas is 13.1%. Now, one is tempted to conclude that since the aforementioned difference of 9.5% is significant, the larger difference of 13.1% is also significant. This would necessarily follow if the results in both instances were based on samples of the same size; but if the samples are not of the same size in both cases, it may well be that the difference of 13.1% is not significant, while the smaller difference is. Such is the case in the ex- ample just cited. STATISTICAL ANALYSIS The purpose of a statistical analysis of this survey is to determine whether a person's living habits have a signifi- cant effeot upon his listening habits. The prdblem is essen- tially one of testing percentages. For example, we may wish to know whether church-goers as a whole tend to listen to classical music more than nonfichurch~goers. we select two random samples, one of church-goers and the other of non-church- goers from the city of Lansing. For the sample of 210 church- goers, 141 (or 67%) listen to classical music; whereas only 105 (or 45.3%)of the sample of 232 non-church—goers listen to classical music. Now, did this difference in percentages occur 14 purely as the result of chance, or did it occur because the percentage of church-goers on the whole who listen to classi- cal music is actually greater than the percentage of non- church-goers who listen to it? Before attempting to answer this question, we must 1. Decide at what level of significance we shall work. 2. Determine the probability that the difference arose from.chance'alone. It was decided to work at the one percent level of sig- nificance as is commonly done. That is, if the probability that a given result is due to chance alone is 0.01 or greater, we shall concede that this result may well be accounted for on the basis of chance alone. But if the probability that a given result is due to chance alone is less than 0.01, we shall not regard this result as something that has occurred by chance. To determine the probability that the difference in per- centages arose purely by chance, we make use of the fact that the sampling distribution of differences of percentages based on large samples is very nearly normal. That is, if we draw repeatedly pairs of random samples of n and n items from two populations in both of which the percentage is2p, and if we compute the difference in sample percentages for each pair of sample, the distribution of these differences in sample per- 15 centageswill be approximately normal (provided n1 and n are large) with mean 0 and standard deviation6p1-p25V n1 + n2 where qanl - p. It is our knowledge of this sampling distri- bution which enables us to make an accurate estimate of the probability that the differences in two sample percentages arose from chance. Just how this can be done can best be ex- plained by means of an example. Consider the aforementioned example involving church’ goers and non-church-goers. Here we have a random sample of 210 church—goers, 67.1% of whom listen to classical music; and a random sample of 232 non-churchrgoers, of whom 45.3% listen to classical music. Thus, n1: 210, p1: 67.1%, n; 233, p 2 45.3%. We wish to determine the probability that this difference in percentages, (namely, 21.8%) resulted from chance. Let us assume that it did; i.e., we assume that-these two samples were taken from two populations, A and B, in which the percentages are the same, say p. Now, we do not know the value of p; we make the best estimate of p on the basis of the sample values, p1 and p . 2 This estimate is n1p1+ nape 141+ 10 “—71—— : ‘fim‘fi— ‘ °~557 1+ 2 Recall that for pairs of samples drawn from A and B, the distribution of percentage differences is very nearly normal 16 ‘ (since n and n are large) with.mean zero and standard 1 2 deviation 6p 10 Vii i E =V(O.§57zgo.443, (0.5512go.4422 = 0,048 1 2" n n 210 232 1 2 Consequently, for any two random samples drawn from A and B, the probability that the difference in sample per- centages, p -p , will be numerically greater than 2.58 p -p1 (20.124) is represented by the areas of the tails of the nay-mil curve in Fig. l and is equal to 0.01, as found from a table of areas for the standard normal curve. fi2.58 “€2.58 (= 3. 124 '. P “P P 'P l 2 l 2 Fig. 1 Now, for the two samples of cl'mrch-goers and non-church-goers, p -p a 0.218: 4.540’p -p . That is, p ~p2) 2.586;: -p , so 1 2 l 2 l l 2 that the probability that this difference in percentages arose from chance alone is much less than 0.01 (actually, the prob- abi li ty 1 s about 0. 000004). 17 we conclude therefore, that the two samples very probably came from populations with different percentages. Or, more specifically, we conclude that the percentage of population A listening to classical music is different from the percentage of population B, where populations A and B consist of all church~ goers and non-church-goers respectively, in the city of Lansing. We are not absolutely certain that our conclusion is correct; we may, in fact, be wrong. But the probability that we are wrong is less than 0.01 (in this case, the probability that we err is about 0.000004). 18 EDUCATION Education appears to have a significant influence upon a person's listening preferences with regard to the follow- ing categories: classical music, dance music, hillbilly, music, drama, discussions and forums. Classical music. The percentage of high school people who listen to class- ical music is not significantly greater than the percentage of grammar school people who listen. However, the percentage of college_peoplewho listen to classical music is signifi- cantly greater‘than the percentage of high school or grammar school people. Dance music. There is no significant difference between the number of grammar school people who listen to dance music and the number of high school people. From our data, we cannot tell definitely whether significantly more grammar school people‘ listen than college people. The observed differences in the said percentages is 17.2; the probability that this differ- once is due to chance alone is almost exactly 0.01. Hence, we cannot really tell whether there is or is not a signifi- cant difference in the two percentages. However, since the '7 People with grammar school, high school, and college educa- tion, respectively. 19 difference in the percentages of high school and college people who listen is significant, one would expect the dif- ferences in the percentages of grammar school and college people to be significant also. But in order to make a def- inite decision, one should repeat the experiment several times. The significant difference occurs between the high school and the college level. It was found that 64.1% of the high school people listen to dance music as compared with 54:6; of the college group. Hillbilly Music. There is no significant difference between the percent- age of grammar school and high school people who listen to hillbilly music. This cannot be said of the high school and the college groups, for it was found that a.gignificantly greater percentage_of high school people listen to hillbilly music than college, and a significantly greater number of grammar school people listen to hillbilly music than college people. Drama. There is no significant difference in listening prefer- ence for drama between either the grammar school and the high school groups or between the high school and college groups. However, the difference between the percentage of grade school and the percentage‘of college people who listen to drama is m ghly si gn‘i r1 cant . 20 This would indicate that there is a gradual increase in pref- erence for'the drama from the grammar school group to the col- legeggroup. Discussions and Forums. There was no significant difference between the prefer- ences of high school and grammar school groups for discussions and forums. likewise, there was no significant difference be- tween the percentage of high school and percentage of college people who listen to discussions and forums. The real differ- ence occurred between the grammar school and college groups; 23.l%'of the former listen to discussions and forums as com- pared with 43.8% of the latter. Again, the growth in prefer- ence for discussions and forums from the grammar school group to the college group was gradual. 21 EDUCATION Tm °F mm (gifii‘fi; 332.33%“ ($2323.35) Classical music 35.9 44.9 63.1 Dance Music 71.8 64.1 54.6 Hillbilly music 28.8 26.3 12.4 Religious Devotions I 41.0 29.5 25.4 mystery Stories 47.4 50.9 -56.9 Comedy 56.4 52.1 56.2 Drama 41.0 52.1 64.6 Soap Operas 34.6 30.8 21.5 Quiz Programs 71.1 58.1 63.8 Discussions a Forums 23.1 31.6 43.8 Shopping Programs 31.5 28.9 18.0 variety Shows 51.6 54.9 51.6 DIFFERENCES IN PERCENTAGES EDUCATION LEVELS (Asterisk indicates significant difference in percentages) TYPE OF PROGRAM Difference Difference Difference ‘ High School High School College Grade School College Grade School Classical music 9.0 18.2* 27.2' Dance music 7.7 9.5' 17.2* Hillbilly Music 2.5 13.9* 16.4* Religious Devotions 11.5 4.1 15.6 Mystery Stories 3.5 6.0 9.5 Comedy 4.3 4.1 0.2 Drama 11.1 12.5 23.6. Soap Operas 3.8 9.3 13.1 Quiz Programs -13.0 5.7 7.3 Discussions a Forums 8.5 12.2 20.7; Shopping Programs 2.6 10.9 13.5 variety Shows 2.8 3.3 0.5 23 INCOME The influence of income upon a person's listening pref- erence is evident in choice of music and religious devotions. Classical Music. In preference for classical music, there is a difference between the lowest income group and the highest, but not be- tween any other categories. There is an appreciable differ- ence here, as is evidenced by the fact that only 42.9% of the low-income group listen to classical music, whereas 71.2% of the highest-income group listen to it. Hillbilly Music. A difference exists between the losest-income group and the highest, with 33.3%{of the lowest-income group listening to hillbilly music as compared to 12.8% of the highest-income group. Again, there is no difference in the other categories. Dance Music. All income groups reacted equally favorably to this type of music. Religious Devotions. The significant difference again appears between the lowest and the highest-income groups, with a gradual declension in preference between the successive groups, so that the highest- income group emerges with the significantly smaller percentage favoring religious devotions. 24 Discussions and Forums. Again, there is a significant difference between the lowest and the highest income groups. The percentage of the highest-income group listening to discussions and forums is significantly greater than the percentage of the lowest- income group. In all other categories, other than choice of music and religious devotions, income range appears to exert little influence of significance. TYPE OF PROGRAM .2000 “3000 3-5000 “15,0003 2 ' Pe ent Percent) Percent) Percent) Classical Music 42.9 57.0 55.3 71.2' Dance Music 51.0 56.0 52.4 57.4 Hillbilly Music 33.3 22.6 24.5 12.8 Religious Devotions 58.8 47.1 ‘ 42.3 35»? Mystery Stories 37.2 51.2 58.3 50.0 Comedy 40. 5 47.7 43. 8 48. 8 Drama 50.0 65.1 56.2 61.3 Soap Operas 26.2 31.4 34.0 18.8 Quiz Programs 61.9 61.6 63.8 65.0 Discussions and Forums 38.1 46.5 38.7 57.! Shopping Programs 33.3 33.3 25.2 16.2 Variety Shows 48.0 50.0 57.5 54.2 Group I GrOup II Group III Group IV $2000 and Under‘ Between 32000 and 3000 and Between 3000 5000 $5000 and Above 26 (Asterisk indicates significant difference in percentages.) TYPE OF PROGRAM Difference Difference Difference Difference in Percent- in Percent- in Percent- in Percent- ages for ages for ages for ages for I;_and I III and II IV'and III IV and I Classical Music 14.1 1.7 15.9 28.3“ Dance Music 5.0 3.6 5.0 6.4 Hillbilly Music 10.7 1.9 11.7 20.5. Religious Devotions 11.7 4.8 6.6 23.1* Mystery Stories 14.0 7.1 8.3 12.8 Comedy 7.2 3.9 5.0 8.3 Drama 15.1 8.9 5.1 11.3 Soap Operas 5.2 2.6 15.2 7.4 Quiz Programs 0.3 2.2 1.2 3.1 Discussions & Forums 8.4 7.8 18.8* '19.4 Shopping Programs 0.0 8.1 8.0 17.1 Variety Shows 2.0 7.5 3.2 6.2 0.7 5.9 3.1 3.1 ‘ Radio Advertising 27 FAMI LY SI ZE The size of the family was found to have no significant influence on the type of program preferred. SIZE OF FAMILX Husband, Husband, Husband, TYPE OF PROGRAM Wife, 2 Wife, 1 Wife or more child children (Percent) (Percent) (Percent) Classical Music 45.6 40.4 42.8 Dmc. MU..1° 6101 56c? 63e6 Hillbilly Music 25.0 19.2 19.5 Religious Devotions 44.6 47.1 41.0 Mystery Stories 55.4 60.6 50.9 Comedy 54.2 61.5 53.8 Dram‘ 4604 54e8 5‘s} Morning or Afternoon 33.3 35.6 34.1 Serials Quiz Programs 71.4 75.0 66.1 Discussions a Forums 35.7 39.4 40.5 Shopping Programs 20.8 26.0 22.5 variety Shows 55.4 64.4 55.5 28 29 Group I - Husband &‘Wife Group II - Husband, Wife a 1 Child Group III - Husband, Wife, 2 or more children Difference Difference Difference in Percent- in Percent- in Percent- TYPE OF PROGRAM ages for ages for ages for II and I III and II III and I (Percent) (Percent) (Percent) Classical Music ' 2.4 15.2 2.8 Dance Music 6.9 4.4 2.5 Religious Devotions 6.1 2.5 3.6 Mystery Stories 9.7 5.2 ' 4.5 Comedy 7e? 7e? oeh' Drama 0.5 8.4 7.9 Soap Operas 1.5 2.3 0.8 Quiz Programs 8.9 3.6 5.3 Discussions e Forums 1.1 3.7 4.8 Shopping Programs 3.5 5.2 1.7 Variety Shows 8., 9.0 0.1 30 OCCUPATIONAL LEVELS listeners were divided occupationally into three cats- gories: Laborers, White Collar.and Professional. The Laborers include all factory workers, artisans of all kinds producing ‘pods either in a factory or private 'enterprise, bus drivers, taxi drivers, tavern keepers, electricians, plumbers, storekeepers, etc. The White Collar refers to salesmen, accountants, business men, etc., where the individual labor is more apt to involve mental skills, personalities, and public rela- tions. The Professional includes the lawyer, doctor, druggist, nurse, underwriter, minister, and teacher. 31 OCCUPATIONAL LEVELS Occupation has a significant influence upon preferr ences for classical music, quiz programs, discussions and forums , and drama. Classical Music. A significantly higher percentage of professional people listen to classical music than White Collar people. The trend toward classical music is gradual from the laboring people to the professional group, to the extent that the differences in percentages for the Laborer and White Collar group is not sig- nificant; whereas the difference in percentages for the Labor- ing and Professional classes is highly significant. Quiz Programs. Again, as we progress from the laboring group to the professional group, the preference for Quiz Programs gradu- ally increases: so gradual is this increase that it is not noticed until we compare the laboring with the professional class. Such a comparison reveals that a significantly greater number of professional people listen to Quiz Programs. Discussions and Forums. With discussions and forums the trend is similar to that of classical music, inasmuch as the percentage difference is 32 noticeable until we compare Professional with White Collar groups. The percentage difference for these two groups is significant, as is also the difference in percentages for laboring and professional groups. Drama. A significantly higher percentage of professional peo- ple listen to drama than white collar people. Again, the trend is gradual from the laboring people to the profession- al group. OC CUPATIQ NAL LEVELS 33 m °F PW P32222212? mazes-AR has... Classical Music 76.8 42.5 35.7 Dance Music 57.1 55.9 55.0 Hillbilly Music 10.7 19.7 21.3 Religious Devotions 42.9 44.1 47.0 Mystery Stories 67.9 58.3 56.2 Comedy 60.7 63.0 53.4 Drama 73.2 46.5 45.0 Soap Operas 25.0 35.4 35.3 Quiz Programs 83.9 70.1 63.9 Discussions & Forums 69.6 37.0 43.4 Shopping Programs 17.9' 31.5 29.3 variety Shows 58.9 58.3 51.8 (Asterisk indicates significant difference in W 34 percentages.) Differences Differences Differences in White 001- in Profes- in Profes- TYPE OF PROGRAM 1ar a Laborer sional and sional and Percentages White Collar Laborer Per- Percentages centages Classical Music 6.8 34.3‘ 41.1* Dance Music 0.9 1.2 2.1 Hillbilly Music 1.6 9.2 10.6 Religious Devotions 2.9 1.2 4.1 Mystery Stories 2.1 9.6 1.7 Comedy 9.6 2.3 7.3 Dram‘ 1e5 2607. 2802* SoaptOperas 0.1 10.4 10.3 Quiz Programs 6.2' 13.8 20.0* Discussions & Forums 6.4 32.6* 26.2‘ Shopping Programs 2.2 13.6 11.4 Variety ShOWS 6e5 006 701 35 RELIGION It would seem that church attendance has little effect upon one's preference for radio programs, with the exception of classical music, religious devotions and quiz programs. It was found that church-going people are more apt to listen to classical music, religious devotions and quiz programs than are non-church-goers. This conclusion was reached after a statistical analysis of the data revealed that the percentage of church-going people who listen to the aforementioned pro- grams was significantly greater than the percentage of non- church-goers. RELIGION N0 CnURCH- TYPE OF PROGRAM CHURCH GOER (Percent) (Percent) Classical Music 45.! 67.1 Dance Music 56.0 54,3 Religious Devotions 36.4 55.6 Mystery Stories 55.6 50.9 Comedy 43.0 48. 8 Drama 56.2 60.3 Soap Operas 33.5 27.6 Quiz Programs 67.8 78.6 \ Discussions a. Forums 1 40.2 45.3 I Shopping Programs 26.5 25.8 Variety Shows 59.6 55.7 37 RELIGION (Asterisk indicates significant difference of percentage) I TYPE OF PROGRAM Difference in percentages of church-goers and non-church- goers. Classical Music 21.8* Dance Music 1.7 Hillbilly Music 3.4 Religious Devotions 19.2" Mystery Stories 4.7 Comedy 5.8 Drama 4.1 Soap Operas 5.9 Quiz Programs 10.8* Discussions 6. Forums 5.1 Shopping Programs 0.7 Variety Shows 3.9 REACTIONS TO RADIO ADVERTISING (Educational Levels) 38 GRAMMAR HIGH SCHOOL COLLEGE (Percent) (Percent) (Percent) LISTENS T0 RADIO 43.0 46.8 32.7 ‘ ADVERTISING BUYING IS INFLUENCED 25.3 27.2 24.7 CAN NAME PRODUCT 16.5 18.3 15.3 A significantly greater percentage of people with high school education listen to radio advertising than college people. In the matter of listening to radio advdrtising there was little difference in percentage between grammar school and high school. No significant difference in percentage was found between the educational levels in buying influenced or the ability to name a product. It is worthy of remark that a very small per- centage of all educational levels is influenced by the radio ad- vertising, according to the listeners' testimony. The actual test was applied when it wasrevealed that a still smaller per- centage of each of the levels was able to actually name a product. 39 REACTIONS TO RADIO ADVERTISING (Income Levels) [$2000 on LESS 32-3000 3-5000 $5000.11)”. ' (Percent) (Percent (Perc.) Percent) LISTEN TO RADIO ADVERTISING 38.1 38.8 31.9 35.0 BUYING IS INFLUENCED 19.0 30.2 17.0 18.8 The percentage of people in the 32-3000 income bracket whose buying is influenced by radio advertising is significantly greater than the percentage of people influenced in the 33-5000 group. This was the only group in which any significant differ- ence was recognized. RADIO DIARY ANALYS§§ Definition of Terms 1. Selective or Not Selective In determining whether a listener is Selective or Not Selective in his tuning habits, the following criteria shall be used: If the person gives evidence of listening to the same station for extended periods of two or three hours, as a regular practice throughout the day and throughout the week as an average, such a person's listening shall be con- sidered "Not Selective.” V However, the person who tunes to a station for several hours, but who does not do this as a consistent practice throughout the week, shall not be considered Not Selective. The person who is Selective in his tuning will turn from a program to the station of his choice, without following the same station for three or four hours. It is anticipated that seldom can one be selective in his choice and focus his at- tention on one radio station for three or four hours at a times 2. Dial Hopping Dial Hopping by definition refers to the practice of moving from one station to another consistently throughout 41 the recorded week's period of listening, remaining with one station not more than a half hour at a time as a usual prac- tice. 3. Regplar‘listeningPattern Regular Listening Pattern refers to a systematized prac- tice of following the same programs, or listening to the same stations at regular hours, consistently throughout the re- corded seven-day period of listening. 4. Trotal Number of Hours The Total Number of Hours may refer to the total hours of radio listening for seven days, as recorded in the Radio Diary. PUEEO DIARY ANALYSIS 42 GRAMMAR HIGH COLLEGE SCHOOL SCHOOL (Percent) (Percent) (Percent) SELECTIVE _ LISTENERS 80.0 51.7 53.8 ”LIONS A LISTENING PATTERN . 27.3 35.2 39.0 DIAL HOPPER -. . 27.8 16.6 34.3 There is no significant difference in percentage between the several educational levels in Selective listening. Though it appears that 80%‘of the grammar school people are selective, it must be pointed out that the sample of the grammar school people was of such size, in comparison with the samples for the high school and college listeners, that statistically there is no significance to this high percentage. 43 RADIO DIARY ANALYSIS AVERAGE NUMBER.OF LISTENING HOURS PER WEEK FOR EDUCATIONAL LEVELS HIGH SCHOOL 34.0 hours per week GRAMMAR SCHOOL 24.0 hours per week COLLEGE 31.8 hours per week AVERAGE FOR COMBINED 32.17 hours per week GROUPS Actually, there is no statistical significance between the average number of listening hours of each of the educa- tional groups. There appears to be a difference of note between the grammar school group and the college group, but it is not significant at the 1% level. If we use the 5% level of Judging this difference, it is barely significant. RADIO DIARY ANALYSIS SUNDAY LI STENING (Mbrning, Afternoon, and Evening) It was found that significantly more people with high school education than grammar school education engage in Sunday morning listening. In the afternoon listening it was evident that sig- nificantly more high school people listen than grammar school, and significantly more college people engage in Sunday afternoon listening than grammar school people. The same high percentage for evening listening was evident among the high school group again. It was found, for instance, that a significantly greater percentage of high school people listen Sunday evening than do grammar school people. Similarly, a significantly greater number of high school people listen Sunday evening than do the college group. There is a correspondingly significant greater number of college people who listen Sunday evening than grammar school people. There was no significant difference in the percent- ages of high school and college people who do not listen on Sundays. However, there was a sharply significant difference between the number of high school people and grammar school people who do not listen Sundays. W Sunday Listening (Morning, Afternoon and Evening) NO SUNDAY IVDRNING AFTERNOON EVENING LISTENING GRAWAR SCHOOL 14.0 11.7 18.7 6.4 HIGH SCHOOL 43.9 A12.1 53.2 19.3 COLLEGE 18.7 30.4» 35.7 14.0 SEASONAL DIFFERENCE IN SUNDAY LISTENING RADIO DIARY ANALYSIS 47 \ NO SUNDAY NDRNING AFTERNOON EVENI NG LI STENTNG November to May 3800 40.9 6403 16e4 June ~ to (October 35.7 40. 4 50. 9 22. 8 Chapter 4. W 48 Chapt or 4} CONCLUSIONS Are people predisposed to listen to certain types of programs because of their manner of life, or is lis- tening purely an individual matter subject to human per- sonality variations as variable, complex, and unpredict- able as human genes? From the study conducted, it would appear that living habits have much less influence upon radio listen- ing than might be presumed. This study revealed that radio listening is influenced by income, education, occu- pation, and church attendance. The effect of income is revealed in one's choice of music. In the number of those who listen to classical music, there is to be found a significant difference be- tween the people of the lowest and the highest incomes, the high9r_pergentage of those listening toclassical music appearing among the highest income people. Similarly, with hillbilly music there is an obseryf h able difference between the lowest and the highest in- come levels. .49 Income_influence may be seen in one's choice of re- ligious devotions. The greater percentage of those listen- ing to religious devotions may be found among those in the lowest income bracket. That is, there are signifi- cantly more people in the lowest income bracket who lis- tééit° religious devotions than in the highest income bracket. In otgegflthan these categories, income influenced listening preference verylittle. I Quiz programs were found to rate very high with ____———+-— all income levels. Hillbilly music, soap operas, and shopping programs rate consistently low with all income groups. It was also found that education exerts an influence upon one 's choice of music, drama, discussions and forums. «H. ,-4—- . m..- .M'— a ’4'- More high school people listen to dance music than col-‘ lege people. Similarly, a greater number of grammar school p people listen to hillbilly music than college people. little difference in the number who listen to serious drama could be observed between the high school and the grammar school levels. Yet there is a recognizable differ- ence in the percentages of grammar school people who listen 50 to drama, and the percentages of college people who listen. Here again, we see a gradual increase in percentage from the grammar school through the high school level, reaching the higher percentage at the college level. This would, of course, indicate that college people, and people of wider education, would be more apt to listen to drama, though serious drama is nonetheless accepted at all levels. In preferenoefor discussions and forums, education playsfapart, apparently, in influencing the college lev- el. Little difference in percentage was found between the grammar school and the high school groups. People's choice for classical music tends to vary with occupational levels. It was found that professional people ”*mw_ prefer classical musip more than do white collar people. iflAM-fi There is to be found a gradual increase in preferencefor classical music, as one proceeds from the laboring group through the white collar group to the professional. While the difference in increased percentages between the sever- al groups is not significant itself, there is at the top a significant difference between the number of profession- al people who listen to classical music and the number of laboring people who listen to it. Likewise, in the choice for quiz programs, the professional group “shows a signi- ficantly greater percentage over the laboring groups of thosewho prefer this type of program. The samc_is true in the matter of listening to discussions and forums. Here again, the professional class prefer this kind of program more than do the laboring class. (It was found that church attendance influences peo- ple_in their choice of classical music, religious devo- / tions and quiz programs. The church attenders showed a marked preference for classical music, religious devotions and quiz programs. While quiz programs are significantly pepular at all levels, it was apparent here that there is an even greater preference for quiz programs among people who regularly attend church than among people we do not. Contrary to expectations,7it was found that the size of awfuamiwly has no significant influence upon the choice of type of programs. ‘7 Matthew N. Chappel and C. E. Hooper, "Radio Audience Measurement”, Stephen Daye, New York, l9A4, p.35. 52 . The public's reaction to radio advertising offered a surprising conclusion. The public at large professes not to take radio advertising too seriously. An average of 33.8% admit listening to radio advertising. Remarks such as, ‘Oh, I listen - - - of course - - - like I lie- ten to the ticking of the clock on the wall; I know it's there," were typical of the answers received, which varied from mild disdain and indignation to tolerant acceptance of what was believed to be necessary in order to pay for the programs that are enjoyed. It was found that 21.2% admitted that their buying had been influenced by radio advertising. It must be remarked here that there is a significantly greater percentage of people whose buying is influenced in the 32000 to 33000 income group than in any other income group. Further, considering the reaction to advertising in the educational levels, it was found that a significantly greater number of people of high school level listen to radio advertising than do college or grammar school people. ‘The analyses of the Radio Diaries evidenced that a high percentage of listeners are Selective. That is, they ”d; not tune to the same station for three or four hours at a time. Similarly, a characteristically low percentage 53 fellow a listening pattern consistently through the week. An average percentage of only 34.1% follow a regular lis- tening pattern. Dial hopping is not too common, except among the college level. Here it was found that a signi- ficantly greater number of college people than high school people are "dial hoppers.“ Sunday listening practices revealed that people of high school level conprise the greater part of the listen- ing audiencefion Sunday morning, afternoon andwevening. It was also shown that a significantly greater number of college people listen Sunday evening than do people with grammar school education. 3‘ Though we may not attach any statistical signifi- cance to it, the Diary nonetheless revealed the fact that people of high school education probably listen the great- er number of hours 913393; Grammar school people listen 24 hours; college people 31.8; high school people, 34. One is ledto conclude that living habits exert lit- tle influence upon listening habits except in the type of _programs that require thought and appreciation based upon training and background. Quiz programs in particular seem to”t§ifiats all these limitations, as do dance music, comedy, 54 variety shows, and drama. Programs featuring hillbilly music, soap operas and shopping programs are unpopular at most levels. It is evident that there are certain programs that provide a common denominator for reaching all levels. Oh; .;n be reasonably assured that quiz programs, dance music, mystery stories, and drama will be accepted at most levels. A The extent to which radio listening is affected by background and manner of living is a question that is very difficult to determine. Outside of the few areas affected by living, indicated in this study, there is the wide general field of human-interest yet to be fully exploited. It remains for the genius to find that common denominator of human appreciation that will extend to all levels of mankind. Certainly, from this study it_may be supposed that programs appealing to the higher (so- calIed) cultural forms of life, programs requiring re- flection and background, programs requiring the effort of thought, are not going to reach the average man. {59; 55 grams for entertainmentflwillfalways be popular; yet it remains for the person devising a program to develop that which interests in an area where thought and appreciation are not the requisite for its enjoyment, but are nonethe-' less to be included in its high edifying quality. Chapter 5 APPENDIX _K r 5785) Michigan State College Dept. of Radio Speech and Dramatics East Lansing Dear Sir or Madam: Mr. Roger Hamlin will call at your home for a brief interview; Mr. Hamlin is a member of the faculty of Written and Spoken English at Michigan State College and is doing research work in the field of radio. Mr. Hamlin's study will require the careful consideration of all of us. Any help by way of information that you may offer him.will be greatly appreciated. Your name has been chosen from some 90,000 in the city of Lansing in an effort to make a cross-section study of the city. The report of Mr. Hamlin's findings will be published. E Cordially yours, 1 germ? Joe A. Callaway Director of Radio Education THE QUESTIONNAIRE When arranging this survey, it wasthougnt advisable to ask a number of questions that mignt offer data for further studies in correlating living habits with radio listening habits. Thus it may be seen that only certain questions referring to education, income, size of family, church attendance and occupation have actually been used in this thesis. It is anticipated that the additional in- formation gained in the questionnaire will offer invalu- able material for further studies. QUESTIONNAIRE This is a radio survey being conducted by Mr. Roger Hamlin of Michigan State College, in collaboration with the Department of Radio, Speech & Dramatics of Michigan State College, and Sponsor Magazine, New York City. Any information that you give will enable us to make a complete and accurate study of this city. Do not feel obligated to sign your name or reveal your identity. 1;:- O\'\n 9. to. 31. Do you have a radio in your car? Do yOu use your radio frequently when driving? Do you tune your car radio to programs regularly? What do you read? Newspapermgwm Magazines? ”"meovel ? mmmthhersmmw m? Does family spend evenings together at home? Last school attended (Grammar, High, etc.) When do you retire? Keep regular hours? Do you attend a church? No WW“,Yesuwmm If so, how often Approximate valuation of property? ‘What do you do on Sundays? Driveumflgm Walk “m“,mAttend movie ,,,,, Read_mwmwm Sleep in...” Church .............. Visit Listen to radio .............. Housework W. Family consists of: Self Husband and Wife Number of children Inlaws What do you do for relaxation when tired? Do you own a car? Kind Year Occupation Male Female What do you ordinarily do evenings? Listen to radio?mmmmmm Newspapers,mwmumwm Cards W ..... Be at W Drive car ._................._ Lodges or clubs...-................r Other .. Estimated income bracket: $2000 $2000 to $3000 $3000 to $5000 ...............Ab°V9 e..."- eoesss - menu’s-s4 I Do you.play golf, ball, bowl, or any other sports? Do you rent a home? Do you live in a rented room? Do you own your own home? 5786) 3786) Do you engage in any community work, such as Parent Teachers, lodges, clubs, etc.? Would you buy a television set if price were lowered within your means?__,,,,_,,___._.,,_,,.,,, Have you ever seen a television show? Do you own a television set? How many people do you know who have television sets? Do you know what "F M" (Frequency Modulation) is? Do you own an "F M" set? ,4“ .W (Check type of program you prefer when you listen to radio) News 1. Sports 2. Newscaster (New reporter) 3. News Commentator (Comments on the news) l. Hillbilly music or cowboy songs 2. Dance music 3. Semi-classicial or classical music h. Symphony orchestras 1. Morning or afternoon serials such as "soap operas" 2. Serious dramas 3. Mystery stories h. Comedy 1. Shopping programs,comments or advice to housewives 2. Quiz programs . variety shows . Discussions and forums such as Town Hall Meeting of the Air, etc. 3 h 5. Religious Devotionals 6 . Others 30 you listen to commerical advertisements on the air? 30 you feel that radio advertising has influenced your buying? If there any product you use or prefer because of radio advertising? If 5°: can you name a few? .Radio .................... Both "mm-non... In. 1W do you prefer to get the day's news? Newspaper? ( 5786) “II'UOIm—OOOOWIOI‘UIAI V—m WJIM ’ ‘m-mwx-w ‘ _ _ L Other. Q15 :30 Q15 :30 215 :30_ WJIM WJR WWJ WKAR Other ' “‘3‘“ W W“ e“ ._.... 4 as" .. . no *’ m: ”c. x 8 3984 I HICHIGQN STQTE UNIV. LIBRRRIES 31293101806325