The writer reviews applications of mathematical models to European marketing problems selected because they foreshadow or complement comparable problems in the United States.
Get full access to this article
View all access options for this article.
References
1.
Jean-MichelAgostini, “How To Estimate Unduplicated Audiences,”Journal of Advertising Research, 1(1961), 11–14.
2.
Agostini has developed a formula which renders unnecessary the laborious tabulation of the unduplicated audiences of combinations of three or more media vehicles. It has been successfully applied to both French and U.S. magazine data (See also [10, 42, 43, 50, 53]).
3.
Jean-MichelAgostini, “Analysis of Magazine Accumulative Audience,”Journal of Advertising Research, 2(1962), 24–27.
4.
Agostini offers another formula, this one for estimating the accumulative audience of several successive issues from just the average audience of one issue and the accumulative audience of two.
5.
AgostiniJean-Michel, “The Case for Direct Questions on Reading Habits,”Journal of Advertising Research, in press.
6.
A sample of 750 housewives chosen at random from a city in northern France were interviewed three times in Spring 1961 on their reading of four widely circulated weekly magazines; 86 percent were interviewed in all three waves. The survey measured the accumulative audience, up to nine issues, for each magazine, by the editorial interest (“through-the-book”) interviewing method. Before this, each housewife was also asked, “Do you read magazine X regularly, occasionally, or never?” Average issue audience as given by the editorial interest method could be accurately estimated from the housewives’ statements on their reading habits given in answer to the latter question.
7.
BaracheM., “Problèmes de Vente et de Distribution: Difficultès Practiques, Rôle de la Recherche Opérationelle dans ce Domaine (Sales and Distribution Problems: Practical Difficulties and the Role of Operational Research in this Field),”Revue de Recherche Opérationelle, 1, 154.
8.
BelsonWilliam A., Studies in Readership, London: Business Publications Limited, 1962, published on behalf of the Institute of Practicioners in Advertising.
9.
In the words of one prominent colleague (See [33]), this is a must for all media researchers. Using intensive interviews by highly trained interviewers as his criterion, Belson finds large errors in respondents’ claims to have read a magazine made upon being shown its flag (name as shown on its cover). In view of the impossibility of “validating” readership claims against actual behavior as observed, this study must be regarded as the most heroic approach to validation yet attempted.
10.
BenjaminB., and MaitlandJ., “Operational Research and Advertising: Some Experiments in the Use of Analogies,”Operational Research Quarterly, 9(1958), 207–217.
11.
Five advertising campaigns (for vitamins, military service, and radio equipment) are described in terms of their inputs (numbers of announcements, journal issues, or leaflets distributed) and outputs (inquiries). Plotted by period, the input-output relationships were fitted by curves suggested by analogy to physiological response to stimuli (logarithmic), the law of diminishing returns (a converging series suggested by Zentler and Ryde), the problem of biological assay—increased dosage will affect what additional percent of the population? (cumulated normal), and the charging of an electrical condenser (exponential). The logarithmic and cumulated normal curves fit the data best.
12.
BenjaminB., and JollyW. P., “Operational Research and Advertising: Theories of Response,”Operations Research Quarterly, 11(1960), 205–218.
13.
Benjamin and Maitland provide further validation of a previously developed analogy between advertising and physiological stimulus and response. A more complex epidemiological model is developed and tested to deal with the distribution of response in time. Jolly investigates the possibility of electronic simulation of advertising response and develops independently a simple exponential expression for coupon response which appears to agree with experimental results and the physiological model.
14.
BremerAxel, “Uber die Anwendung der Vektoren Analyse in der Okonomischen Verhaltensforschung (On the Application of Vector Analysis in the Study of Economic Behavior),”Metro, 1(1962), 205–220.
15.
A consumer's behavior involves a conflict between his desire for the product and his resistance to paying its cost. The author finds deficiencies in Bilkey's idea that the pattern of consumer behavior is equal to the sum of many independent results. By understanding how Bilkey departs from the original ideas of Kurt Lewin, it is possible to use the limited but valuable opportunities offered by Lewin's vector analysis in the study of economic behavior.
16.
CaffynJ. M., EhrenbergA. S. C., MonkD. M., and NuttallC. G. F., The Audience of Television Advertisements, London: The London Press Exchange Limited, 1962.
17.
This is a more detailed presentation of the work described by C. G. F. Nuttall in “TV Commercial Audiences in the United Kingdom” (See [55]).
18.
CaffynJohn M., and SagovskyMarianna, “Net Audiences of British Newspapers: A Comparison of the Agostini and Sainsbury Methods,”Journal of Advertising Research, 3(1963), 21–25.
19.
A 25-year-old technique for obtaining net audience gives accurate estimates when duplicated audience is high.
20.
CalviG., “La Differenziasione Marginale dei Prodotti et il Comportamento di Scelta del Consumatore (Marginal Differences Between Products and Consumer Choice Behavior),”Archives of Psychology, Neurology and Psychiatry, 22(1961), 111–121.
21.
The buyer must choose among similar products. To avoid making a difficult choice he often focuses on marginal characteristics of the product, e.g., its wrapping, as the differentiating element.
22.
CohenJ., and CooperP. A., “A Study of a National Lottery: Premium Savings Bonds,”Occupational Psychology, 34(1960), 170–183.
23.
This English study found purchasers of National Lottery Bonds in higher occupational groups much less optimistic about winning, but owning considerably more bonds. More of them than of the lower group felt that winning a prize would yield more than saving.
24.
CoplandBrian D., “Exposure and Communication Measures of Outdoor Advertising in Britain,”Journal of Advertising Research, 1(1961), 13–17.
25.
Two British studies of outdoor advertising used different criteria of effectiveness. The author discusses the pros and cons of each criterion and its application to advertising in other media.
26.
CoplandBrian D., “Some Fundamentals of Poster Audience Measurement,”Journal of Advertising Research, 2 (1962), 20–27.
27.
The audience of an outdoor display appears to be directly proportional to the number of passages by it, be it large or small, in town or city, in Europe or the U.S.A.
28.
CoplandBrian D., A Review of Poster Research, London: Business Publications Ltd., 1963, published on behalf of The Institute of Practicioners in Advertising and the British Poster Association.
29.
The definitive work in a small domain where quantitative regularities are happily the rule rather than the exception.
30.
DaleniusTore, “Treatment of the Non-Response Problem,”Journal of Advertising Research, 1(1961), 1–7.
31.
“Survey estimates will never be unbiased—too many in the sample can't be reached.” Dr. Dalenius deplores such defeatism and suggests three ways of handling non-response.
32.
Renéde Chantal, “L'Action des Sciences sur la Gestion Future (The Influence of Science on Tomorrow's Managers),”Travail et Méthodes and L'étude du Travail, Paris, 1955.
33.
DeLuciaL., Problems of Sampling Techniques in Market Research—Estimation of Proportions, Quaderni dell'Associazione di Mercato, Rome, 1958.
34.
A collection of lecture notes from a course in market research, on the estimation of proportions in random sampling of attributes.
35.
DreyfusH., and MontyG., “L'Approvisionnement, les Marches et la Recherche Opérationnelle (Supply, Markets and Operational Research),”Revue de Recherche Opérationnelle, 9(1958), 171–195.
36.
These experiments with new procurement schedules for flooring of railway carriages by the French National Railways were based on analysis of bids (including timing and price-volume relations) furnished by 800 regular suppliers, and the volume of deliveries required during the month after the orders were placed.
37.
DutertreRoger, “Une Règle Simple de Gestion des Fabrications à Court Terme (A Simple Rule for Short-term Production),”Revue de Recherche Opérationnelle, 5(1962), 313–332.
38.
Seasonal demand fluctuations may be satisfied either by stockpiling or by changing production rate. The best compromise between these two solutions is studied, each one receiving an estimated cost. The management system may be considered as an automatic mechanism in discrete variables with two inputs (real demand and difference between it and forecasts) and two outputs (growth of production and of stock related to forecasts).
39.
EhrenbergA. S. C., “The Pattern of Consumer Purchases,”Applied Statistics, 8(1959), 26–41.
40.
A definite pattern exists in the proportions of people who buy various quantities of a product during a fixed time period. Different products and periods give rise to different curves, but Mr. Ehrenberg shows that for a wide variety of products and with only mild restrictions on the period, the curves are all members of the negative binomial family.
41.
EhrenbergA. S. C., “A Study of Some Potential Biases in the Operation of a Consumer Panel,”Applied Statistics, 9(1960), 20–27.
42.
Length of time the respondent has been in the panel appears to have no significant influence on reported brand shares or reported frequency of purchase. Mr. Ehrenberg admits a possible “warm-up” bias for new panel members, though he has no evidence to present one way or the other on this subject. The second source of bias, the amount of information collected from panel members, seems equally nonsignificant, at least for respondents who have been reporting for several months and who are asked to report on 50 product groups instead of their usual 30.
43.
EhrenbergA. S. C., “How Reliable Is Aided Recall of TV Viewing?”Journal of Advertising Research, 1(1961), 29–31.
44.
Claims of viewing television were virtually identical for recall periods of from one to seven days.
45.
EhrenbergA. S. C., “Verified Predictions of Consumer Purchasing Patterns,”Commentary, 10(1963), 16–21.
46.
Two new applications of his earlier work (see above) show how consumer purchasing data tend to follow a negative binomial distribution. First, a simple approximate formula for the standard error of the average quantity bought is given. Apart from the potential usefulness of this formula for everyday application, the formula has been used to establish the influence—which turns out to be drastic— of the detailed pattern of purchasing on the size of sampling errors (see p. 19). Second, the proportion of buyers in a period longer than that actually observed can be successfully predicted, if there is no trend in the market. From there, predictions of “loyal” buyers and the like can also be deduced.
47.
FabensA. J., “The Solution of Queueing and Inventory Models by Semi-Markoff Processes,”Journal of the Royal Statistical Society, 23(1961), 113–127.
48.
A semi-Markoff process (one where intervals between transitions are distributed randomly) is used to determine the limiting distributions of two queueing systems. In the first, the customers arrive at random and are served in batches of K, the service-times having a general distribution. At the completion of a service period the customers served leave the queue and another service period starts if there are at least K waiting; otherwise, the server rests until there are K customers present. In the second, the customers arrive singly with inter-arrival time having a gamma distribution of order K and are served singly. If times between purchases and delays in the production of new stock are randomly distributed, and if customer requests may be backlogged if there is no stock, the process behaves exactly as does a Poisson-input, batch-service queue.
49.
FouilhéP., “évaluation Subjective des Prix (Subjective Estimates of Prices),”Revue Française Sociologique, 1(1960), 163–172.
50.
Consumer attitudes toward prices, and their estimation, could be expressed in terms of a logarithmic scale. This tends to confirm Stevens’ power law that equal sensation ratios correspond to equal response ratios.
51.
GaborAndre, and GrangerG. W. J., “On the Price Consciousness of Consumers,”Applied Statistics, 9(1961), 170–188.
52.
The authors explored the extent to which 428 Nottingham housewives were aware of prices paid for 15 frequently purchased grocery articles. Prices were correctly quoted for 57 percent of all purchases reported. Of those considered incorrect, half deviated from the listed price by less than 10 percent. By commodity, tea was the winner with 79 percent of tea purchases correctly priced by the housewife.
53.
GennaroPietro, “Le Modalita Di Esposizione e L'Efficacia Delia Vetrina (Method of Display and the Effectiveness of Show Windows),”Rivista Internazionale di Scienze Economiche et Commerciali, 5(1958), 3–15.
54.
The problem is how to use store window space to the best advantage. A model to measure the duration of the window exposure to the passing customer shows that even a slight difference in the left hand—right hand proportion of traffic before the window results in wide differences in the exposure value of the different parts of the window space. An actual count of traffic in two commercial streets in Milan, at four places, confirms differences in traffic composition.
55.
GennaroPietro, “Distribuzione Ottimale del Bilancio Publicitario per Veicoli e Nel Tempo (Optimal Distribution of Advertising by Vehicle and Time Period),”Quaderni dell'Associazione di Mercato, Rome, 1960.
56.
The author suggests that time scheduling of advertising effort, a neglected subject, may be one of the most important determinants of advertising efficiency. He theorizes that level of advertising awareness in a given section of the public reached by the specific campaign at a given moment may be interpreted as a combined effect: (a) of the conversion function, which produces a growing awareness with the increase of the number of impressions according to a logarithmic scale; (b) of the number of dropouts depending, according to an hyperbolic function, on the time passed since the last impression was received; and (c) of the number of customers recovered from the dropouts when advertising appears again. He tries to determine the three functions using data collected by Zielske and shows how the media schedule might be rationally set, taking into account the net margin per product unit, the cost of each advertising series, and the three functions.
57.
GennaroPietro, “Un Problème de Promotion: Distribution du Budget Publicitaire sur les Différents Supports en Concurrence (A Promotion Problem: Allocation of the Advertising Budget among Different Media Used at Once),” chapter 31 in Application des Méthodes Modernes, Paris, 1962. Friedman's model for geographical allocation of advertising expenditures is adapted to allocation of impressions to different media, taking into account the relative strengths of competitive efforts. Applied to the allocation of Esso advertising in five Italian newspapers, the model's schedule differs from that actually used.
58.
GiardiniBasilio, “Casi di Applicazione de Ricerca Operativa (Examples of Applications of Operations Research),”La Scuola in Azione, 10(1962), 5–170.
59.
This is a comprehensive survey article, including a general introduction to O.R. methods (choice of alternatives, construction of models, decision criteria, etc.), plus numerous examples of linear programing, dynamic programing, production and inventory control, and queues.
60.
GuettaPierre, “L'Utilisation des Mésures d'Attitude dans le Prétest ou le Controle de l'Éfficacité de la Publicité (The Use of Attitude Measurement in Pre-testing or the Control of the Effectiveness of Advertising),”Metra, 1(1962), 415–426.
61.
The effects of advertising are usually uncontrollable once sales are under way because of distribution, competition, economic climate, etc. The ability to recall an argument is no guarantee that it has been accepted; on the contrary, this ability may well prevent any attitude change whenever the source of the message is not considered reliable or disinterested. Instead the author took as a criterion the changes of attitude observed in a population under laboratory conditions, using a questionnaire before and after advertising was shown, and using sequential sampling surveys to uncover changes in attitude of the population.
62.
HenonR., “Prix de Revient et Politiques de Ventes (Net Cost and Market Planning),”Revue de Recherche Opérationnelle, 1, 104–105.
63.
HenryHarry, “Belson's Studies in Readership,”Journal of Advertising Research, 2(1962), 9–14.
64.
A researcher now in management, Mr. Henry advises: “… get hold of this book, read and digest it, then go back and start learning the techniques of questionnaire design and field control right from scratch.”
65.
HeynWolfgang, “Problems of the Calculation of Significance in Market Research,”Unternehmensforschung, 3(1959), 130–136.
66.
HessEva–Marie, “Methodische Uberlegungen bei der Anlage von Leserschraftsuntersuchungen fur Periodisch Erscheinede Aeitschriften (Methodological Reflections on Readership Studies of Periodicals),”Metra, 1, 271–280.
67.
In the German Federal Republic, 12,000 individual oral interviews annually determine coverage of 45 magazines with more than regional distribution. Because users of the analysis demand equal treatment of all magazines in the study, methodological analyses were done to see if any biases had occurred. The error resulting from non-response varied by magazine. A clustering effect was observed for magazines with strong market concentrations. Readership analyses, “especially because of their thematic center of gravity,” thus cannot be carried out with the aid of quota-questions. Seasonal fluctuations must be taken into account and interviews evenly distributed over survey weeks and survey days.
68.
Institute of Practitioners in Advertising, Operational Research: A Report, I.P.A. Occasional Paper No. 13, London: Institute of Practitioners in Advertising, 1961, prepared and presented by D. Lowe Watson.
69.
This is a nontechnical review of the nature of operations research and its applications to advertising. Several examples are discussed, including the models developed by Benjamin, Jolly and Maitland, and Lee and Burkart. The most useful areas for application appear to be determination of the optimum size of the advertising budget, and allocation of the budget among media and over time.
70.
JeanneauP., “La Prévision par Marche des Ventes de Gaz à Usage Domestique (Forecasting by Sales Price of Gas for Domestic Use),”Revue de Statistique Appliquée, 10(1962), 85–97.
71.
A sample inquiry inventoried equipment of subscribers to gas in the suburbs of Paris from 1950 to 1956. The author estimated probabilities of transition from one level to another between the two periods. Sales forecasts assumed that consumption did not change within each level.
72.
KemsleyW. F. F., “Designing a Budget Survey,”Applied Statistics, 8(1959), 114–123.
73.
The author discusses the design of a survey to obtain information on consumer expenditures. Since most families do not keep complete budgets, he recommends the use of a diary by a sample of households. The article describes some of the difficulties in obtaining accurate reports from the diary method and lists six areas the field worker should check for accuracy.
74.
KentA. J., and NuttallC., “An Attempt to Answer the Question—Does Advertising Pay?”The Incorporated Statistician, 10(1960), 29–35.
75.
The authors express a measure of the sales of two competing British candy products by the current trend of candy consumption, cumulative advertising expenditure on each of the two products, free spending money available, the relative price of the two products, advertising expenditures on all products of the same kind, and seasonal effects. After discarding some of these variables as inconvenient or irrelevant, they conclude that advertising and sales are related but that the relationship should be interpreted with great caution, if at all.
76.
KlanferJ., “Théorie de l'Information et Publicité (Information Theory and Advertising),”Revue de Recherche Opérationnelle, 6(1962), 35–54.
77.
Advertising is a type of information which must operate through filters (customers). The quantity, I, of information transmitted can be quantified in bits as in information theory and depends on the number of classes of customers, n, on their proportions, pi, on special aspects of the advertising created by the advertisers, on the volume of advertising, and on the percent of receivers in each population, ni as follows:
78.
The author discusses the negative influence of redundancy and advertising “noise.”
79.
KrewerasG., “Stratégie de Vente Sur un Marche de Durée Limitée (Sales Strategy in a Market of Limited Duration),”Revue de Recherche Opérationnelle, 1, 252.
80.
KuhnWalther, “Die Nettoreichweitenbestimmung Einer Kombination von Zeitschriften Ein Naherungslosung (Net Readership Calculation of Combinations of Magazines),”Metro, 1(1962), 153–164.
81.
In order to determine the net readership of a combination of more than two magazines, one needs only the individual readership figures for each magazine and the overlap of each pair of magazines. German experience found the relationship exponential: e-kx, where k is a coefficient depending upon the number of magazines. The correspondence between calculated and real numbers of readers was satisfactory; the largest deviations were about 1.8 percent.
82.
KuhnWalther, “Net Audiences of German Magazines: A New Formula,”Journal of Advertising Research, 3(1963), 30–33.
83.
The equation one should use to estimate the net audience of a combination of vehicles depends on the number of vehicles. This is an abbreviated translation of the preceding article.
84.
LambertP., “Étude de la Prévision des Ventes Pour Articles de Grande Consommation (Study of Sales Forecasting for Mass Consumption Products),”Revue de Statistique Appliqueée, 6(1958), 59–80.
85.
This article studies a mathematical model of sales of high consumption articles which is intended to explain the irregularities of sales and the consumer reactions which influence sales. Stocks of finished products play the part of buffers which enable production to be flexible; however, inventories are immobilized stock and should be minimized.
86.
LecchiSandro, “La Scelta Degli Ispettori delle Vendite: Un Applicazione della Teordella Code (Selection of Sales Inspectors: An Application of the Waiting Line Theory),”Bolletino del Centro per la Ricerca Operative, 4(1960), 14–28.
87.
Lecchi discusses different aspects of business organization and solves a problem of selecting sales inspectors with a decision model using available sales data.
88.
LeeA. M., and BurkartA. J., “Some Optimization Problems in Advertising Media Planning,”Operational Research Quarterly, 11(1960), 113–122.
89.
Here is described the first stage of a multi-stage research program in the planning of advertising media schedules. The problems of securing maximum impact and maximum coverage in an intense campaign of short duration for a fixed outlay are discussed. A precise solution of the impact problem and an approximate solution of the coverage problem are outlined.
90.
LeeA. M., “Decision Rules for Media Scheduling: Static Campaigns,”Operational Research Quarterly, 13(1962), 229–242.
91.
Lee states the media scheduling problem and presents an advertising model from which come decision rules applicable to static campaigns, i.e., those intended to evoke a maximum response, the time of attainment being unimportant. (See also Lee's “Letter to the Editor” in vol. 14 of ORQ, p. 89, in which he corrects two errors in the above model and defends its assumption that the probabilities of reading different vehicles are independent.)
Commercial research is defined as “scientific preparation for commercial decisions.” Four main functions of commercial research are suggested: market research, explanation of commercial phenomena, forecasting, and experimentation.
94.
LongtonPeter A., and WarnerBernard T., “A Mathematical Model for Marketing,”Metra, 1, 297–310.
95.
Behavior of customers choosing among alternative brands was described by a matrix of transition probabilities. A formula was developed to predict brand shares from brand loyalty and the pressure to buy a brand exerted by external forces such as advertising. The relative growth coefficient for sales of one brand over one cycle of the process was considered, and a formula developed for the special case where all loyalties are equal and the market is in equilibrium. Company growth was related to purchasing pressure expressed as a ratio of its share of the market, and evidence presented for the empirical validity of this relationship. This “Dynamic Difference Formula” is a special case of the more general formula for relative growth. The probability distribution of market share was formulated by means of a model borrowed from population genetics. Concepts include systematic drift due, for example, to an advertising campaign, and stochastic drift due to other effects acting on a group of finite size.
96.
MarcMarcel, “Net Audiences of French Business Papers: Agostini's Formula Applied to Special Markets,”Journal of Advertising Research, 3(1963), 26–29.
97.
Monsieur Marc extends his colleague's method to the specialized audiences of French business papers.
98.
McAnallyPatrick, “Market Research and O.R.,”Operational Research Quarterly, 2(1951).
99.
The idea of market research as a branch of O.R. is considered by examining several recent publications. Differences appear. The end product of market research is often publication, which differs from facilitating particular decisions. Preoccupation with sampling survey techniques seems foreign to O.R., which is interested in what happens before and after. There also appear differences in presentation of results as facts or as recommendations for action.
100.
MeleseJacques, “L'Application des Méthodes de Recherche Opérationnelle à des Problèmes de Distribution et de Vente (Application of the Methods of Operational Research to Some Problems of Distribution and Sales),”Collège de Recherche Opérationnelle des 25 et 26 Nov. 1956, Organisé par l'Institut d'Administration des Entreprises de l'Université d'Aix-Marseille.
101.
The first study sought to determine the number and location of factories and warehouses to meet future plans for a food products company. The solution, based on topological notions, led to the development of a set of theorems for locating towns for warehouses, for separating a territory into independent zones, and for fixing the optimum location within the zones by an iterative method. The second study determined the interior layout of a store by means of factorial experiments designed to maximize total sales of all lines.
102.
MetheringhamR. A., “Adding a New Dimension to Media Evaluation,”Journal of Advertising Research, in press.
103.
This method not only yields net accumulative audience of a combination of vehicles from their duplications in pairs, but also estimates the frequency distribution of the advertising exposures of a print media schedule from the duplication between different issues of the same publication.
104.
MoonenH. J. M., “Het Bepalen van Destelniveau's Wanneer Afname en Levertijd Gamma-Verdeeld Resp. Normaal Verdeeld Zijn (Determining the Reorder Level If the Demand Is Gamma-Distributed and the Delivery Has a Normal Distribution),”Statistica Neerlandica, 16(1962), 113–120.
105.
If demand is gamma-distributed and delivery time is normally distributed, the probability of running out of stock for a given reorder level can be represented by an integrand, which is, however, not integrable. Rules are given for computing the reorder level for a permitted probability of out-of-stock. A simplified model, in which the demand during the (variable) delivery time is also taken to be gamma-distributed, gives results which differ little from those obtained with the exact method.
106.
NuttallC. G. F., “TV Commercial Audiences in the United Kingdom,”Journal of Advertising Research, 3(1962), 19–28.
107.
When the commercial starts, viewer attention wanders. New measures, reliable and sensitive, show how far.
108.
PompiljGiuseppe, “La R.O. nelle Imprese Industriali e Commerciali, nelle Aziende Agricole e in Altri Enti (O.R. in Manufacturing and Commercial Enterprises, Agricultural Concerns, and Other Institutions),”La Scuola in Azione, 9(1962), 66–78.
109.
Here is a discussion of the nature of O.R., the application of O.R. methods to various types of organizations, and a brief survey of some O.R. activities in Italy.
110.
RhenmanEric, “The Organization as a Control System,”Ekonomisk Tidskrift, 64(1962).
111.
The organization can be regarded as a behavior system with productive components, which are involved in the process of transforming raw materials into products, and administrative components, which are only taking part in the processing of information. The purpose of the administrative subsystem is to control the production process. Control depends on exchange of information. Further development of the model and factors limiting control are discussed.
112.
RosenfeldFelix, and SalomonMichel, “Utilisation de Modèles Markoviens et Pseudo-Markoviens Dans les Études de Marche (The Use of Markov and Pseudo-Markov Models in Market Surveys),”Metro, 1(1962), 221–229.
113.
Markov processes or pseudo-Markov processes can be used to analyze and represent the growth of demand for durable goods or of products used in their manufacture, when the units of consumption are distributed in distinct classes. Demand for domestic coal in France was studied. Units of consumption were classed into six levels of average consumption of coal. Because data referred to two separate periods and was collected for other purposes, it was difficult to determine the probability of movement. Acceptable solutions were achieved by using certain working hypotheses.
114.
SauermannHeinz, and SeltenReinhard, “Adaptation Theory of the Firm,”Zietschrift fur die Gesamte Staatsweissenschaft, 118(1962).
115.
The authors propose a new approach to the problem of entrepreneurial behavior, based on Simon's concept of bounded rationality. Described are a model of routine behavior and a model of planning. Both make use of two analytical instruments: the adaptation scheme (Anspruchanpassungsschema) and the influence scheme (Einflus-schema). The adaptation scheme reflects the preference system, and the influence scheme is a rough picture of the environment. The adaptation scheme attaches “urgency orders” and “resign variables” to the grid points of a rectangular grid in the goal space formed by the goal variables. The grid points are potential aspiration levels. Adaptations to a higher level of aspiration follow from the urgency order while adaptations to a lower level of aspiration are always made with regard to the resign variable. The influence scheme is a matrix indicating negative, neutral, or positive influences of actions (for example: raising advertising expenditures by five percent) on goal variables by -, 0, +, respectively. If the influence scheme is not confirmed by experience, it is corrected accordingly.
116.
StoetzelJean, “A Factor Analysis of the Liquor Preferences of French Consumers,”Journal of Advertising Research, 1(1960), 7–11.
117.
If complex behavior is caused by a few simple motives, they can be found in the patterns of that behavior. Professor Stoetzel illustrates this premise with a factor analysis which tentatively explains consumer preference for nine liquors by their sweetness, price, and regional popularity.
118.
Van Den DriesscheRene, “Markt-Segmentation und Mehrfach-Korrelationen (Market Segmentation and Multiple Correlation),”Metro, 1(1962), 467–472.
119.
W. A. Belson has suggested a simple way to analyze surveys of mass consumption goods: by those factors which show the highest correlation with the intensity of the studied consumption activity. This method gives more useful results than the usual multiple correlation techniques. Chocolate consumption in Germany is cited as an example.
120.
Van OmmeslagheBernard, “Analyse de la Prospection Optimum d'Une Société Commerciale de Distribution (Optimal Canvassing Policy for a Distributing Company),”Metro, 1(1962), 281–296.
121.
It was possible to calculate the number of sales representatives a trading company should employ to maximize profit. The function maximized was the difference between the company's gross margin and the sum of the salaries of the sales representatives, car expenses, costs of warehousing, insurance and handling of goods, costs associated with invoicing, costs of delivery by truck, and fixed administration expenses. By an analysis of the company's cost and by surveys, this was expressed as a function of the number of retailers canvassed, and then maximized graphically. Thus it was possible to go from the number of retailers to be canvassed to the optimal number of sales representatives to be employed.
122.
VenturaE., “Comment Calculer un Barème Digressif, un Système de Ristournes (How to Compute a Table of Scale-Variable Discounts),”Revue de Recherche Opérationnelle, 5(1962), 333–344.
123.
Price elasticity was deemed to be too crude to solve this problem. The present model takes account of the percentage of actual customers who buy presently a quantity x at price to, and are able to buy a quantity y (y > x) at a price t(t < to). The problem is to find, with to given, the quantity y* and price t* to be applied for y* > y so that the total profit is at a maximum. The model was applied to the pricing of food products and required less than one month to work.
124.
VishnevskiiA., “Basic Principles and Methods of Determining the Economic Efficiency of New Technical Measures in Communications Economics,”Vestnik Svyazi, 17(1957). An English translation may be purchased from the Office of Technical Services, Department of Commerce, Washington 25, D. C.
125.
Zora TerolProcopio, “A Sequential Decision Problem in Economics,”Trabajos de Estadistica, 9 (1958), 103–110.
