Innovative Health Care Alternatives: a Model of the Consumer Behavior Process

Stephen J. Miller, Oklahoma State University
William G. Zikmund, Oklahoma State University
ABSTRACT - The choice among health care alternatives by the individual is a consumer decision process. The authors integrate prior health care research into a schematic model of the decision process. Then they empirically examine the model constructs as explanation of intentions to patronize a nurse-practitioner.
[ to cite ]:
Stephen J. Miller and William G. Zikmund (1979) ,"Innovative Health Care Alternatives: a Model of the Consumer Behavior Process", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 628-634.

Advances in Consumer Research Volume 6, 1979      Pages 628-634

INNOVATIVE HEALTH CARE ALTERNATIVES: A MODEL OF THE CONSUMER BEHAVIOR PROCESS

Stephen J. Miller, Oklahoma State University

William G. Zikmund, Oklahoma State University

ABSTRACT -

The choice among health care alternatives by the individual is a consumer decision process. The authors integrate prior health care research into a schematic model of the decision process. Then they empirically examine the model constructs as explanation of intentions to patronize a nurse-practitioner.

INTRODUCTION

Our nation is increasingly becoming aware of the need for improved health behavior and health care. To this end, solutions of a "production" orientation have emerged with emphasis on financial support for more doctors, nurses, hospitals, and insurance plans. Yet there exist continuing problems of inadequate health care, particularly among disadvantaged population segments such as the poor, ethnic minorities, the lower educated, and the elderly.

Long range solutions will likely require innovative delivery methods. The most publicized of these are the nurse-practitioner and physician-assistant programs. Other delivery components might be the mobile medical units of a "circuit rider" nature, day clinics and emergency helicopter service. While the medical profession may view these activities as accepted delivery system components, to the typical user/consumer, the physician is the primary entrance mechanism to the health care system. Alternatives often are perceived as innovations fraught with uncertainties, risks, and sundry misgivings. The development of such delivery system alternatives will be no means assure their acceptance by the public.

Only recently have social scientists come to realize the need for a "health care consumer" orientation (e.g., Aday and Andersen 1975; Andreopoulus 1974; Freeman, Levine, and Reeder 1972) as integral to improving the health of the public. Evidence has begun to accumulate indicating that social-psychological factors influence the range of decision involved in preventive, diagnostic, and therapeutic care. Therefore, the behavioral forces at work on the demand side of the health care equation must be given serious attention.

RESEARCH OBJECTIVES

Rural America represents a prime focus of attention for disadvantaged consumers regarding access to health care (Andreopoulus 1975; Luft, Hershey and Morrell 1976). Disproportionate numbers of poor, elderly, and low educated individuals comprise the population. Likewise, distances of 30-40 miles to health care personnel and facilities, antiquated facilities, and "overworked" elderly physicians are common occurrences. Yet, the rural consumers may very well reject the innovative health care delivery system necessary to assure their access to primary care. Attention must turn to the psychological factors that influence their acceptance or rejection of such offerings.

This research reports on a study of the choice regarding the decision to patronize a nurse-practitioner located in a rural community for diagnostic or therapeutic care versus the alternatives of physician (or perhaps direct avoidance of medical aid). The research objectives that guided the study were:

1. How receptive are the individuals to a nurse-practitioner in the community?

2. What socio-economic and psychological factors explain variability in willingness to patronize a nurse-practitioner?

3. How do the factors that enter the patronage intention decision differ by sex?

More specifically, the research examines a number of constructs felt to be relevant components of the decision. The model uses constructs from the field of medical "sociology" and theoretical constructs in consumer behavior. (The work of Suchman 1966; Rosenstock 1966; Zaltman and Vertinsky 1971; and Howard and Sheth 1969 are the prime base for the model).

Although health care decision making can be viewed in the traditional consumer behavior literature, the operational concepts in brand choice situations have different meanings in their usage across specific situations. Thus, generalized expectancies should not necessarily be formulated at the highest level of abstraction (e.g., personality). Instead, domain-specific expectancies such as "medical fatalism" can be more fruitfully inter-grated into the health care model. Exploratory development of measures for such domain-specific constructs have been utilized in the research.

The section to follow will examine the literature germane to this research and the relevant constructs to be examined. Then the empirical measurements will be explained. Finally, the results of the study are presented.

A MODEL OF HEALTH CARE UTILIZATION

A great deal of research has been conducted relative to utilization of health services. Selected reviews of the more important research are presented by Aday (1972) and by Andersen and Bartkus (1973). Based on this past research, the health care model for the research has been developed. Figure A schematically portrays the variables in this model. The model focuses on a decision regarding a health care innovation. Thus, it is oriented to the fundamental decision whether or not to utilize the innovation. The proximal variables are hypothetical constructs within the domain of the specific decision process. The distal variables are general characteristics and predispositions that provide the context within which health care decisions are made. Of course, there is a specificity-generality continuum along which all the variables lie. Proximal variables will be considered first in the model discussion.

The decision process in the medical care choice will be activated by receipt of stimuli. A decision maker might receive a variety of stimuli that activate decision making. Any given stimulus might be a physiological state such as pain, a social comment such as "you look pale today" or a myriad of other stimuli.

FIGURE A

Receptivity to the stimuli is a function of the psychological readiness on the part of the decision maker. Given readiness, the evaluation of alternative responses will follow. The evaluation will include a combination of perceived benefits (attitudes) and risks. If uncertainties exist, search for information will occur. Likewise, a variety of inhibitors (barriers) exist that will mediate the attitude role as it influences the behavioral intention. The key components of the evaluation process are discussed below.

Symptom Sensitivity

The immediacy and strength of response to the stimulus will relate to psychological readiness. Readiness has been defined as a two-component dimension by Rosenstock (1966) that reflects sensitivity to health-based stimuli as a mediating factor. It is a composite of perceived susceptibility to the given condition and seriousness of the condition. The susceptibility component is the subjective probability of illness contraction as perceived by the individual. The seriousness dimension is the perceived consequence, either medical, economic, or social. For example, an individual may view cancer as a low probability disease with a very serious consequence. In contrast, a headache may appear moderately likely with little concern for its seriousness.

The psychological readiness construct has less specificity than others considered as proximal variables. However, it is discussed at this point because of its expected initial role in the decision process.

Health Care Attitudes

A decision to act given receipt of a stimulus primarily focuses on the perceived benefits of taking such action (c.f., Rosenstock, 1966). Such perceived benefits have been particularly valuable in the study of preventive health behavior. For example, a relationship between perceived benefits and behavior has been noted for chest X-rays (Hochbaum, 1958), dental check-ups (Kegeles, 1963) and inoculations (Leventhal et al., 1960).

The benefit attitudes are by nature domain-specific. In this research instance, they will deal with beliefs regarding the effectiveness of utilizing the health service innovation for treatment of a perceived illness. Benefits in the case of nurse-practitioners might include items such as reduction in waiting time, lower travel or fee costs, greater empathy, and relative medical competency.

Using a factor analysis, 12 specific beliefs toward nurse-practitioners were reduced to three main dimensions or benefit factors. These are competency of the nurse-practitioner to perform the task, interpersonal relations with the patient, and comparative performance of the task relative to the more traditional physician (Zikmund and Miller, 1977).

Perceived Risk

In conjunction with perceived benefits, the decision maker will likely consider the possible negative dimensions of treatment by a nurse-practitioner, treatment risk. This construct has strong empirical validation in the consumer behavior literature. First introduced by Bauer (1960), he remarked that "consumer behavior involves risk in the sense that any action of a consumer will predict consequences which he cannot anticipate with anything approximating certainty, and some of which, at least, are likely to be unpleasant." Thus, a consumer's perceived risk is composed of (1) the subjective uncertainty as to the outcome of actions and (2) the possible consequences related to any given action. Subsequent investigations have viewed perceived risk as multi-dimensional phenomena (e.g., Zikmund and Scott, 1973).

Four major types of risk will most likely apply to nurse-practitioner consideration by the individual. It is believed that there will be a financial risk of utilizing a nurse practitioner. That is, it may not be worth the money to visit a nurse-practitioner or it may be a waste of money. There may be a social risk in the sense that a person's friend or significant others' appraisal of a nurse practitioner may be perceived to be negative (Franklin and McLemore, 1970). There is further a performance risk in the sense that a nurse practitioner may not perform as skillfully as a medical doctor and perhaps may cause pain to the individual. Finally, there is a health risk in the sense that the person's condition may remain the same or even worsen. Other forms of risk such as convenience risk may be important in the sense that there may be a time loss by visiting a nurse practitioner. However, these dimensions of risk are believed to be of less importance.

Inhibiting Factors

A wide variety of inhibiting factors might be considered. These could be included as either proximal or distal to the decision process depending on how situation-specific the variables are defined.

The ready availability of services influences utilization of health services (Bice and White, 1969). Of particular relevance in a model of nurse practitioner selection is the factor of distance, since the nurse-practitioner study focuses on rural areas. The distance component would include both physician and hospital dimensions of availability.

The method by which one pays for medical care is an inhibitor of action. Income is discussed later as a distal variable. However, directly health-linked payment methods can be viewed as inhibitors. For example, Klarman (1963) and Andersen and Sheatsley (1959) both found lower usage rates among prepaid group practice plans than among fee for service plans.

GENERAL INFLUENCES

The distal variables are those of a general nature that influence the decision process as structural factors. Their influence is expected to be less direct than that exerted by the proximal variables. The major categories of distal variables are discussed below.

Socio-demographic Influences

A number of variables related to socio-economic statue have been found relevant to health care decision making. Suchman in a series of articles (1966a; 1966b) investigated ethnic and social factors in medical orientation. He found that social group structure, i.e., cosmopolitan versus parochial groups differed in medical orientation. He also found that certain demographic factors such as age, sex, and social class influenced medical orientation, knowledge about disease, skepticism of medical care, and dependency in illness. Further, demographic factors, medical orientation, and health status influence the source of medical care (1966b). Subsequent research supports the influence of socio-economic status variables (c.f., Greersen et al., 1975).

In a behavioral model of health services developed by Andersen (1968), predisposing variables that established a health care orientation and need were identified. These included socio-demographic and social-psychological correlates. Finally, age is an important predictor of health utilization, and the relationship between physician utilization and age tends to be U-shaped, i.e., old people and very young people tend to use more services. Education increases the probability of overall physician usage, particularly with respect to preventive health care behavior. Marital status, family size, race and ethnicity have also shown some influences on physician utilization. Of particular significance in understanding health care decision processes is the likely structural differences in the decision process by sex. It has been noted that men and women encounter the need for health services at different stages in their lives (Booth and Babchuck, 1972). Also, the woman usually plays the health care decision maker for the family as a whole, particularly for herself and the children.

Under the label of enabling factors as posed by Andersen (1968), economic status has shown significant relationships to physician utilization. For example, the higher the income, the greater the utilization of physician services, Price of medical services and methods of financing are related to income and appear to be important factors.

Health Orientation

The individual possesses a number of general opinions concerning health and health care matters. These are drawn from past experience and general social influences. Suchman (1966a) has found medical skepticism to play an influence on the health care decision. The skepticism construct reflects an uncertainty as to beneficial returns from the medical profession and health care behavior. Suchman also found a psychological dependency in illness construct to reflect a health orientation. In past studies, these have not been extremely valuable in predicting the use of health services (Aday, 1972, p. 21). However, they are included in the current analysis since the consumer decision being investigated is an innovative choice.

A health orientation variable that appears relevant to the nurse practitioner decision is medical fatalism. This construct is based on Zaltman and Vertinsky's (1971) fatalism and is in line with Rotter's (1966) notion of internal-external control of reinforcement. It is believed that certain individuals will hold that being in good health depends more on being lucky, chance, etc., than on any activities within the control of the individual. The more fatalistic an individual is, the less likely he is to see any benefits resulting from his or an agency's efforts to improve his health status.

The health history should have a variety of influences on the health care decision, This is partially felt through the effect of health history on the health orientation as learned responses. However, direct influences may well occur (Bice and White, 1969; Wirick, 1966; Andersen and Bartkus, 1973). One would expect that a history of frequent physical exams, a family doctor, and other links to an established delivery system would negate receptivity to alternative delivery systems. On the other hand, a frequency of past illnesses would likely stimulate general interest in an expanded system.

RESEARCH METHOD

The research data for examining the model construct of FIGURE A were obtained through in-home personal interviews with a cross-section of adults residing in small Oklahoma communities. These rural areas include concentrations of poor, elderly, and lower educated citizens. Thus, an environment exists that is conducive to alternative health care delivery systems.

In conjunction with representatives from the Oklahoma Health Planning Commission staff, a judgment sample of 10 communities in Oklahoma were selected for study. Criteria in the selection were: communities of approximately 2500 residents or less; diverse health care delivery system capacities over communities (ranging from no physician-hospital to two physicians-hospital); geographic and socio-economic diversity over communities. Visual surveys of housing types and conditions were conducted within each community then households were selected for inclusion in the study on a purposive basis to yield socio-economic diversity, In total, 220 adult household members were interviewed. Women were sampled more heavily since they have traditionally played the greater role in household health care decisions. From the original sample, 205 usable questionnaires were obtained (148 women, 57 men).

The data collection was be a personal interview with a variety of question formats, both open and closed ended and required approximately 45 minutes to administer. Pretesting of the questionnaire for clarity and ease of administration was conducted in a non-surveyed community.

The measurements for model constructs are given in FIGURE B. Most are standard measures (e.g., Medical Skepticism, Psychological Readiness, Dependency in Illness, Health History and Demographics). Others specific to nurse-practitioners (e.g., benefits and risks) and the Medical Fatalism scale have been constructed for this research.

Prior to the questions linking directly to nurse practitioners, the respondents were read a description that explained the training of such medical personnel, their general treatment activities and a plan for the nurse practitioner to maintain a primary care office in the community while working closely with a physician in a nearby town for referral and advice.

RESULTS

The respondents were queried regarding familiarity with the nurse practitioner concept along with two other innovative delivery methods to disguise the study's purpose. Familiarity and general orientation toward the medical innovation are shown in TABLE 1. There was little initial familiarity. Forty-five percent of the respondents had never heard of a nurse practitioner while another 31 percent possessed only a basic recognition. Thus, only 24 percent had any explicit cognitions linked to the concept.

Willingness to visit a nurse practitioner was considerably more favorable than the familiarity status. Note that 73 percent express some degree of willingness while only 17 percent reject the concept.

The "willingness to visit" is general. As such, it does not measure the intensity of patronage intention over a variety of situations. Patronage intent was measured based on anticipated response to an array of fifteen symptoms from minor to major significance (e.g., fall, chest pains, high blood pressure). The intention results are presented in TABLE 2 for both men and women. These indicate the average number of medical symptoms for which the individual expressed an intention to visit a nurse practitioner for treatment should one be available.

TABLE 1

RECEPTIVITY TO NURSE PRACTITIONER

For the fifteen symptoms included in the query, the average number for which a respondent would patronize a nurse-practitioner was slightly more than 50 percent for women and less than 50 percent for men. The differences in patronage intention for women and men were statistically significant at the .001 level with the results indicating less receptivity to nurse practitioners by men than by women. This should likely be interpreted as a relative perspective since women tend to be more habitual health care users than men due to pregnancy, pediatric care for children, menopause, and other health care needs throughout their lives.

FIGURE B

MODEL VARIABLES

TABLE 2

PATRONAGE INTENTION

STRUCTURAL CORRELATES

Each variable in the model was correlated with patronage intention. These correlations are given in TABLE 3. Of the forty-eight correlations, nineteen are statistically significant at the .1 level or beyond. Most are from the proximal set of values.

All correlations for the benefit attitude dimensions are statistically significant for both men and women. The correlations are positive indicating attitude-intention relationships in the expected direction. They are somewhat stronger for the men than for women.

Perceived risk plays a key role in patronage intention consideration with three of the risk components being statistically significant for women and for men. As noted by the negative correlations, the higher the perceived risk, the lower is the intention to patronize a nurse-practitioner. Interestingly, the statistically significant risk components differ by sex. Social risk is a salient criteria in consideration for women while the performance risk correlation is statistically significant only for men. Financial and health risks are statistically significant correlates for both sexes.

TABLE 3

PATRONAGE INTENTION CORRELATES

Among the socio-demographic variables, only age among women is statistically significant. Younger women appear more receptive to nurse-practitioners than do older women. This result is congruent with the traditional innovation literature.

Only one statistically significant relationship emerges among the health orientation concepts. This is medical fatalism. Among men, the degree of internal control correlates positively with patronage intentions.

Although none of the health history constructs are significant at the .1 level, for men the variables (especially family doctor) all point toward receptivity to this delivery system mechanism. Since men have far fewer contacts with health personnel than women (either alone or with children), the recent incident of exposure to illness and treatment may modestly increase receptivity to and trust of health care innovations.

A number of interesting relationships emerge for the inhibiting factors. For men, the availability of local health care personnel and facilities lowers the patronage intention. When contrasted to the insignificant relationships among women, this suggests a convenience orientation for men. Yet women, with perhaps an orientation toward continued medical relationships and a concern with the role of household health care coordinator, may find this factor of less importance in the consideration of alternative health care delivery methods. The physical distance to the more traditional physician seems to favor local patronage of nurse-practitioners for both sexes although it is statistically significant only for women. Finally, the availability of medical insurance relates to participation by women. Perhaps this relationship partially explains the weaker correlation for financial risk among women than among men.

The distal variables exhibit far fewer significant correlations than do the proximal variables. Since they have been frequently found to be explanatory in other health care research, it is likely that they play a moderating role in a decision process perspective taken in the current research.

REGRESSION MODELS

Separate regression models were developed for both women and men using intention as the criterion variable and the model variables as predictors. This form of analysis suffers from the assumption of direct, additive, and independent relationships among the variables involved. However, at this exploratory stage of decision model development it has major advantages. First, the cumulative influence of the predictor variables can be assessed. Second, the appearance of variables in the model that had insignificant bivariate correlations can provide clues to underlying conditional variables. Finally, should the relative influence of variables on the decision differ by sex, it will be indicated.

The regression models are given in TABLE 4. They were developed by stepwise regression methods with inclusion of variables terminated where the beta coefficients were not significant at the .1 level. The variables are illustrated by their order of entry and cumulative contribution to explained variation.

A number of observations are in order with regard to the regression models. First, both models are statistically significant (p < .0001) and explain a moderate proportion of the variation in patronage intention. For men, the coefficient of determination, R2, is .44, while for women the R2 is .24. It appears likely that the difference in explained variation can be attributed to the greater complexity in the women's decision, she undoubtedly evaluated the choice with a mediating force of family influences in mind.

The comparison of variables that appeared in the regression models for men and women is informative. As an overview, both proximal and distal variables entered the regressions. For men, variables appeared for attitude, risk and inhibitor constructs emerged from the analysis.

TABLE 4

PATRONAGE INTENTION MODELS: STEPWISE REGRESSION

In both regression models illustrated in TABLE 4, the proximal variables dominate the explanation. The salient attitude and risk constructs combine with the family doctor variable as sources of 70% of the explained variation for men. In the case of women, the salient attitude and risk constructs combine with distance to a physician to yield 75% of the explained variation.

A striking result in comparing the regressions for men versus women is that no specific construct is common to both groups. For men, performance risk is explanatory while for women the key risk variable is social risk. Likewise, with regard to attitudes, competency beliefs were dominant for men while performance beliefs appeared for women. Finally, the dominant distal variables for men are predisposition factors (children, education, illness) while inhibitors are relevant for women (distance to health care personnel, availability of local technology - facilities and insurance).

The variables that entered the regression models differed somewhat from those statistically significant through bivariate correlations. Naturally, a few of the variables with significant bivariate correlations didn't enter the equation due to moderate intercorrelations among the predictor variables. Also, a few variables likely entered the regression as a reflection of conditional explanatory power. For women, this included the local availability of physicians and hospitals. For men, children, family doctor and education emerged in the regression models.

CONCLUSIONS

The major concern of the research was to investigate rural citizens' acceptance of nurse practitioners as an alternative to the present health care delivery system. A model was developed to determine the health care utilization characteristics likely to predispose one to accept nurse-practitioners as the principal primary care delivery system mechanism.

This research has shown that individuals living in rural communities are receptive to the nurse practitioner concept. The receptivity to this innovation in the health delivery system does vary within the population. The correlation analysis indicates that proximal cognitive factors (benefit attitudes and perceived risks) are quite helpful in the prediction of willingness to patronize a nurse practitioner. Thus, it appears the level of specificity of independent variables is extremely important in the prediction of patronage intent.

The multiple regression model provides an empirical assessment of the health care model with the simple framework of linear additive assumptions. The empirical analysis produces explained variances of 44% for men and 24 % for women. These are reasonably high R2 values considering the exploratory nature of this research. It is interesting to note that determinant predictor variables differ by sex. For men, predictive efficiency may be greater than women ostensibly because of males' lower propensity to seek treatment of any form. Thus, men may have a simpler cognitive structure regarding health care innovations (and health care in general).

The model posited and the findings are exploratory. Yet they do merge the traditional consumer behavior theory with research in the health care field. This should serve to stimulate further consumer behavior research in the health care field and provide directions for fruitful efforts.

REFERENCES

L. A. Aday, The Utilization of Health Services and Indices and Correlates (Washington, D.C.: National Center for Health Services Research and Development, DHEW, Publication No. HSM 73-3003, 1972)

L. A. Aday, Development of Indices of Access to Medical Care (Ann Arbor, MI: Health Administration Press, 1975)

R. Andersen, A Behavioral Model of Families' Use of Health Services. Research Series, No. 25 (Chicago: Center for Health Administration Studies, University of Chicago, 1968)

James G. Andersen and David E. Bartkus, "choice of Medical Care: A Behavioral Model of Health and Illness Behavior," Journal of Health and Social Behavior, 14 (1973), 348-62.

Odin W. Anderson and Paul B. Sheatsley, "Comprehensive Medical Insurance--A Study of Costs, Use and Attitudes Under Two Plans," Research Series No. 9, (Chicago: Center for Health Administration Studies, University of Chicago, 1959)

Spyros Andreopoulus, Primary Care: Where Medicine Fails (New York: John Wiley & Sons, 1975)

R. A. Bauer, "Consumer Behavior vs. Risk Taking," Dynamic Marketing for a Changing World (Chicago: American Marketing Association, 1960)

Thomas W. Bice and Ken L. White, "Factors Related to the Use of Health Services: An International Comparative Study," Medical Care, 10(1969), 261-71.

Alan Booth and Nicholas Babchuck, "Seeking Health Care from New Resources," Journal of Health and Social Behavior, 13(1972), 90-99.

Billy Joe Franklin and Dale S. McLemore, "Factors Affecting the Choice of Medical Care among University Students" Journal of Health and Social Behavior, 2(1970), 311-319.

Howard E. Freeman, Sol Levine, and Leo G. Reeder, Handbook of Medical Sociology, Second Ed.,(Englewood Cliffs, NJ: Prentice-Hall, Inc., 1972)

R. Greersen, "A Reexamination of Suchman's Views on Social Factors in Health Care Utilization," Journal of Health and Social Behavior, 16(1975), 226-230.

Godfrey M. Hochbaum, Public Participation in Medical Screening Programs: A Sociopsychological Study (Washington: Public Health Service, Public Health Service Publication No. 572, United States Government Printing Office, (1958)

John A. Howard and Jagdish N. Sheth The Theory of Buyer Behavior(New York: John Wiley & Sons, Inc., 1969)

Stephen S. Kegeles, "Some Motives for Seeking Preventive Dental Care," Journal of the American Dental Association, 67(1963), 90-98.

H. E. Klarman, "Effects of Prepaid Group Practice on Hospital Use," Public Health Reports, 17(1963), 955-65.

Howard Leventhal, et al., "Epidemic Impact on the General Population in Two Cities," The Impact of Asian Influenza on Community Life: A Study in Five Cities (Washington: United States Department of HEW, Public Health Service, Publication No. 766, 1960)

Harold S. Luft, John C. Hershey, and Joan Morrell, "Factors Affecting the Use of Physician Services in a Rural Community," American Journal of Public Health, 66(1976), 865-871.

I. Rosenstock, "Why People Use Health Services," Milbank Memorial Fund Quarterly, 14(1966), 94-127.

J. B. Rotter, "Generalized Expectancies for Internal Versus External Control of Reinforcement," Psychological Monographs, 80(1966)

Edward A. Suchman, "Ethnic and Social Factors in Medical Care Orientation," Milbank Memorial Fund Quarterly, 14 (1966a), 69-78.

Edward A. Suchman, "Social Patterns of Illness and Medical Care," Milbank Memorial Fund Quarterly, 14(1966b), 69-78.

Grover C. Wirick, Jr., "A Multiple Equation Model of Demand for Health Care," Health Services Research, 1 (1966), 301-46.

Gerald Zaltman and Ilan Vertinsky, "Health Services Marketing: A Suggested Model," Journal of Marketing, 35 (1971), 19-27.

W. G. Zikmund and Stephen J. Miller, "Rural Household Attitudes Toward Nurse Practitioners: A Factor Analytic Evaluation," Faculty Working Paper, College of Business Administration, Oklahoma State University, 1977.

W. G. Zikmund and Jerome E. Scott, "A Multivariate Analysis of the Perceived Risk, Self-Confidence and Information Services," Advances in Consumer Research, 1(1975)

----------------------------------------