Provider-patient communication traditionally has been characterized as an asymmetric relationship in which providers exert more power and control over the interaction in comparison to patients (Ballard-Reisch, 1990; Brashers, Haas, & Neidig, 1999). In recent years, however, health communication scholars, medical researchers, and health care practitioners have examined the benefits of patient empowerment through the practice of collaborative communication between providers and patients. This approach involves open discussion between patients and providers about health concerns and opportunities for mutual problem solving and decision making when discussing patient cases and treatment options (Balint & Shelton, 1996; Levinson et al., 1999; Young & Flower, 2002). Collaborative communication has been linked to greater satisfaction with medical encounters, better adherence to treatments, and positive health outcomes, such as reduced stress, pain, and symptom distress (see Brown, Stewart, & Ryan, 2003; Golin, DiMatteo, & Gelberg, 1996; Greenfield, Kaplan, & Ware, 1985).
In order for these benefits to accrue, people must take a more active role and engage in self-advocacy during interactions with their health-care providers. Patient self-advocacy can be defined as representing one's own interests within the healthcare decision-making process (Brashers et al., 1999; Brashers & Klingle, 1992). According to Brashers et al. (1999) patient self-advocacy demands three characteristics on the part of those seeking health care. First, people must have relevant medical knowledge to participate in making decisions about their health. Second, people need to be assertive to a certain degree with their health care providers in terms of initiating topics, seeking clarification, and questioning suggested treatment recommendations. Third, people must sometimes be willing to challenge their health care providers when they feel that the treatments suggested are insufficient or incompatible with their needs.
Patient self-advocacy rests on the assumption that people are willing to discuss health issues. However, as in other contexts of communication, where people vary in their willingness to communicate about topics in general (see McCroskey & Richmond, 1998), there is evidence that individuals vary in their willingness to communicate about health issues (see Brashers et al., 1999; Gotcher & Edwards, 1990; Morgan & Miller, 2002; Smith, Kopfman, Massi-Lindsey, Yoo, & Morrison, 2004). Moreover, physicians often ignore their patients' questions and concerns, forcing patients to be even more assertive if they intend to meet their needs in these situations (Beisecker, 1990). In such situations, those who are unwilling to communicate with their health care providers may suffer more.
In addition, there are many other situations where people communicate about their health with other providers and medical staff, family and friends, and other people with similar health care needs, such as in support groups (Smith et al., 2004; Wright & Frey, in press). These relationships are also important sources of health information for people, yet health issues are often difficult topics for people to discuss due to the social stigma attached to some diseases and conditions (e.g., HIV, AIDS, mental illness, and alcoholism) and the potentially embarrassing information about one's body that can come up in conversations about health, among other reasons (see DiMatteo & Hays, 1981; Gotcher & Edwards, 1990).
Given the importance of patient empowerment and self-advocacy, there is a need to explore underlying traits, such as willingness to communicate about health. Given the significance of the communication trait of willingness to communicate in other contexts and within the health care context, in particular, the purpose of this study was to develop a measurement instrument to assess people's willingness to communicate about their health and to see how this communication predisposition might be related to health information-seeking behaviors and patient assertiveness.
We first review literature supporting the concept of willingness to communicate about health as a potential personality trait and studies that explored the more general trait of willingness to communicate within health contexts. We then discuss the development of the willingness to communicate about health scale and report the findings from a study designed to assess the reliability and validity of this measure.
Willingness to Communicate about Health as a Potential Personality Trait
Although there has been much debate about whether personality predispositions should be conceptualized as enduring ways of being or the enactment of a way of being in particular situations (a "state" approach), however conceived, there is general agreement that traits reference a generative mechanism that predisposes people to behave, to think, or to feel in particular ways (see Beatty, 1998). As Daly and Diesel (1992) explained, personality traits are "directly relevant to communication insofar as each is related to how people communicate or how they come to understand others' messages" (p. 405).
More recently communication scholars have focused on communication traits, those traits that "account for enduring consistencies and differences in individual message-sending and message-receiving behaviors among individuals" (Infante, Rancer,& Womak, 2003, p. 77). Such communication traits include, among others, aggressive communication, communication apprehension, and willingness to communicate (for a recent review of communication traits in the group context, see Keyton & Frey, 2002).
The communication trait of willingness to communicate was defined by McCroskey and Richmond (1998) as "an individual's predisposition to initiate communication with others" (p. 120). Evolving from work conducted by Burgoon (1976) on "unwillingness to communicate," Mortensen, Arnston, and Lustig (1977) on predispositions toward verbal behavior, and McCroskey and Richmond (1982) on shyness (see McCroskey & Richmond, 1987), the communication trait of willingness to communication was first introduced by McCroskey and Richmond in 1985, followed by the construction of an instrument to measure it (McCroskey & Richmond, 1987), and the demonstration of the reliability and validity of the willingness to communicate scale (McCroskey, 1992).
Although McCroskey and Richmond (1987) viewed willingness to communicate as cutting across situations, they also argued from the start that this communication trait was "probably to a major (though as yet undetermined) degree situationally dependent" (p. 129). Accordingly, a number of scholars have studied willingness to communicate within specific situations, showing it to be an important predisposition in a variety of situations, including instructional (Menzel & Carrell, 1999), family (e.g., Avtgis, 1999; Rocca & Martin, 1998), organizational (Richmond & Roach, 1992), and intercultural communication (Lin & Rancer, 2003; McCroskey & Richmond, 1990) settings. In a number of these cases, scholars have created context-specific willingness to communication instruments to measure this communication trait.
Although the communication trait of willingness to communicate seems highly relevant to health care, only three studies have examined it in that context. Two of those studies concern people's willingness to communication about organ donation, with Morgan and Miller (2002) examining the effects of knowledge, attitudes, and values on willingness to communicate about organ donation to family members, and Smith et al. (2004) investigating relationships between willingness to communicate about organ donation within families and variables such as knowledge about organ donation, attitudes towards organ donation, and altruism. The third study by Crowell (2004) compared, in addition to other variables, differences between individuals who were HIV positive and HIV negative with regard to their willingness to communicate about condom usage.
The Crowell (2004) study did not attempt to construct an instrument per se to measure this specific health-related aspect of willingness to communicate (only a single item). Morgan and Miller (2002) and Smith et al. (2004) both employed a willingness to communicate about organ donation scale developed by Morgan and Miller (2002), with all of the items on that scale specifically mentioning organ donation. Although these studies are certainly relevant to the communication trait of willingness to communicate about health, these studies were not conducted for the purpose of constructing a willingness to communicate scale that could be used in a wide variety of health-related situations and interactions.
This lack of developing a willingness to communicate about health scale that could be used in a wide variety of health-related contexts and interactions is unfortunate, for one would expect that willingness to communicate about one's health, in general, would significantly affect people's approach to and interactions in health care settings and with health care personnel. Given the importance of willingness to communicate as a general communication trait, and empirical evidence that suggests people vary in their comfort levels in terms of discussing their health, we sought to develop a generalized measure of willingness to communicate about health that would be potentially useful for both health communication researchers and health care practitioners. After constructing this instrument based on a review of the relevant literature, we assessed the underlying dimensions of this measure, its reliability, and its relationship to other conceptually related variables, in a study of two different populations. The following sections report the development of this instrument and the findings associated with its testing, followed by a discussion of the nature and use of this instrument.
Data for this study were collected from two distinct samples of participants who were asked to complete a questionnaire containing the willingness to communicate about health scale, created by the first two authors, and related measures.
The first sample was from a larger study of patient communication, and it consisted of 95 people living with cancer selected from a cancer center in a metropolitan city in the Southern United States. The average age of the sample was 59.81 (SD = 12.64). Thirty of the respondents were men and 65 were women. The majority of the sample was Caucasian, but 21% of the participants were African Americans. Clients ranged in their affiliation with the cancer center from three visits to 25 years. The following types of cancer and number of people with it were represented in the sample: breast cancer (26), lymphoma (17), colon (10), lung (9), prostate (4), leukemia (4), and other (25).
Given the relatively small size of the first sample, and the desire to assess how the scale might differ for individuals facing more generalized health concerns (as opposed to an illness like cancer), we assessed the willingness to communicate about health scale using a second sample. Undergraduate students enrolled in communication courses at a midsized university in the Southern United States participated for extra course credit. We decided to use undergraduate students for the second sample because such students constitute a significant portion of the general population and are likely to be more similar to that larger population in terms of everyday health concerns than individuals facing cancer. Two hundred and thirty-one students volunteered to participate in the study, of which 139 were women and 92 men, with a mean age of 28.78 years (SD = 10.64). One hundred and twenty-nine self-identified as white, 87 as African Americans, and the rest as Asians, Latina/Latino, or other. The average number of annual physician visits for this sample was 7.04 (SD = 7.83).
Willingness to communicate about health scale (WTCH)
On the basis of relevant health communication and willingness to communicate literature, the first two authors created an instrument to measure cancer center clients' willingness to communicate about their health. This instrument employed a 5-point Likert scale (5 = Strongly Agree, 1 = Strongly Disagree) to evaluate 10 items measuring a person's willingness to communicating about his or her health with a variety of people, including health care providers and nonproviders (see Table 1). The factors discovered for this scale are discussed in the results section. The scores for the items for each factor were summed and the arithmetic mean of the summed scores constituted the operationalization of participants' willingness to communicate about their health, with higher scores indicating more willingness to communicate.
Willingness to communicate scale
The Willingness to Communicate Scale was developed by McCroskey and Richmond (1987). According to McCroskey (1992), this 20-item instrument (with eight filler items) was designed as a direct measure of a respondent's predisposition toward approaching or avoiding the initiation of communication. Participants indicate the percentage of time (from 0 to 100%) they would choose to initiate communication in a variety of situations. The instrument includes three subscores based on the type of person with whom one interacts (strangers, acquaintances, and friends), and four subscores based on type of context in which communication takes place (public, meeting, group, and dyad). This instrument was administered to participants in the second sample only, for the purpose of seeing its relationship to the WTCH, and demonstrated an overall Cronbach's alpha reliability coefficient of .91.
Patient self-advocacy scale (PSAS)
This 12-item instrument developed by Brashers et al. (1999) measures three dimensions of patient communication: education, assertiveness, and nonadherence. Two dimensions of this scale were of particular theoretical interest to the present study for several reasons. First, the PSAS items measuring the education dimension reference an individual's desire for health information, and include items such as "I actively seek out information on my illness." Theoretically, a person who is interested in seeking information about his or her illness should be more willing to communicate about his or her health. In terms of the assertiveness dimension, which includes items such as "I frequently make suggestions to my physician about my health care needs," it makes conceptual sense that a patient who is more assertive would be more willing to communicate about his or her health. It is less clear how the third dimension, nonadherence, might be conceptually related to an individual's willingness to communicate. Although one item, "If my physician prescribes something I don't understand or agree with, I question it," specifically mentions a communicative act initiated by a person. This instrument was administered to the second sample only and demonstrated an overall Cronbach's alpha reliability coefficient of .70.
Exploratory Factor Analysis
For both the cancer patient and the undergraduate student samples, an identical version of the 10-item WTCH scale was used. In both samples, the responses to the scale were subjected to an exploratory factor analysis. The principal axis factoring method first was used to extract the factors, with no restrictions placed on the number of factors extracted. Factors were selected based on two criteria that indicate the largest factors in the data set: an eigenvalue greater than 1.00 and "above the elbow breakpoint" rule of the scree test. The oblimin method then was used for oblique rotation of the factor structure. This method allows for correlated factors, which was likely in the case of willingness to communicate about health matters. In interpreting the rotated factor pattern, an item was said to load on a given factor if the factor loading was .35 or greater for that factor and was less than .35 for the others. After the final factor structure was determined, the reliability of each factor was examined using Cronbach's alpha coefficient.
Cancer patient sample
The factor analysis of the WTCH scale for the cancer patient sample resulted in three factors with eigenvalues greater than 1.00. These values and the corresponding percentages of variance explained were 3.46 (34.67%), 1.37 (13.73%), and 1.08 (10.76%), with the cumulative contribution of the three factors 59.16%. The scree test, however, suggested two clear factors. An oblique rotation was conducted for these factors, which converged in nine iterations. The rotated factor structure and the standardized factor loadings for the items are shown in Table 1. The three factors, in the order of percentage of variance explained, were labeled as willingness to communicate about health with providers (Factor One), willingness to communicate about health with nonproviders (Factor Two), and willingness to seek health information/discuss health issues (Factor Three).
The third factor was comprised of only scale item 1, and scale item 2 exhibited a substantial cross-loading across factors. Hence, the factor analysis was rerun after dropping these two items and the results revealed a two-factor structure (see Table 1). The eigenvalues for the two factors were 2.84 and 1.21, explaining 35.49% and 16.32% of the variance, for a cumulative contribution of 51.81%. The scree test also showed two factors.
In addition to looking at the factor structure of the WTCH in the full student sample, we also decided to examine the structure in two different subgroups of students: those with a low number of reported physician visits versus those with a high number of reported physician visits. Given that the health status of cancer patients differs considerably from typical undergraduate students, we sought to create a student subgroup (i.e., those reporting high physician visits during the past three years) that would be somewhat comparable to the cancer patient sample. Thus, an analysis was conducted for the full sample, and then a separate analysis was conducted for the two subgroups created out of the number of reported physician visits. The two physician visit subgroups were formed after conducting a median split on the number of physician visits in the last three years (M = 7.04; [M.sub.d] = 6; SD = 7.83), resulting in 0-6 visits (n = 158; M = 3.27; SD = 2.01) as the low physician visit group and 7-50 visits (n = 72; M = 15.33; SD = 9.37) as the high physician visit group. The difference in number of physician visits between the two groups was statistically significant, t(228) = 12.21, p < .001.
A listwise deletion procedure for missing data reduced the sample size for some of the analyses to 230, with 138 women and 92 men. The factor analysis of the WTCH scale for the full student sample resulted in three factors with eigenvalues greater than 1.00. These values and the corresponding percentages of variance explained were 3.25 (32.46%), 1.25 (12.50%), and 1.15 (11.49%), for a cumulative contribution of 56.45% for these three factors. The scree test, however, pointed to only two clear factors. An oblique rotation was conducted for these factors, which converged in eight iterations. The rotated factor structure and the standardized factor loadings for the items are shown in Table 2. The three factors, in the order of percentage of variance explained, were labeled, as before, as willingness to communicate about health with providers (Factor One), willingness to communicate with about health with nonproviders (Factor Two), and willingness to seek health information/discuss health issues (Factor Three). There were no items that exhibited a low loading on a factor or substantial cross-loading across factors.
The factor analysis for the low physician visit subgroup showed a pattern similar to the full sample. After factor extraction and rotation, three factors that were similar to the full sample emerged (see Table 3), and the scree test also showed three factors. Their relative contribution to explaining the variance in the data set was 31.27% (eigenvalue = 3.13), 13.39% (eigenvalue = 1.34), and 12.73% (eigenvalue = 1.27), respectively, with the cumulative percentage equaling 57.39%. Unlike the full sample, Factor Three explained a greater percentage of the variance than did Factor Two. Similar to the full sample, no items showed cross-loadings between factors, although scale item 4 showed low loading on its factor.
The factor analysis for the high physician visit subgroup also exhibited a similar structure overall, but a somewhat different pattern of factor loadings. After factor extraction and rotation, three factors also emerged. The scree test, however, showed only two factors, with their relative contribution to explaining the variance being 48.14%(eigenvalue = 1.41), and 10.37% (eigenvalue = 1.04), respectively, for a cumulative total of 58.51%. However, these factors were dissimilar to the full sample and the low physician visit subgroup. As Table 3 shows, Factor One was comprised of scale items 6, 7, 8, 9, and 10, which was similar to the full sample and the low physician visit subgroup. However, scale item 5 loaded on this factor and not on Factor Three as in earlier analyses. Furthermore, scale item 3 was the only item that loaded on Factor Two, and item 2, which loaded on this factor in earlier analyses, now loaded on Factor Three. Thus, for the high physician visit subgroup, the first factor was comprised of scale items 5, 6, 7, 8, 9, and 10; the second factor of only item 3; and the third factor of items 1, 2, and 4.
Correlations among factors
The correlations among the three factors for the cancer patient sample with all items ranged from .19 to .42. The correlation between the two factors for the items-deleted analysis was also in the same range (see Table 1). The three factors for the full student sample and low physician visit subgroup were comprised of similar items and exhibited moderate correlations, ranging from .23 to .49 (see Tables 2 and 3). The correlations among the three factors for the high physician visit subgroup were -.19, -.41, and .32, respectively. Thus, the three factors were sufficiently independent of each other.
The items corresponding to a particular factor constituted a subscale. The measurement reliability of the three subscales was assessed using Cronbach's alpha coefficient, with a coefficient of .70 and above indicating sufficient reliability. Coefficients for the three factors for the cancer patient sample, full student sample, and low physician visit student sample ranged from .60 to .71 (see Tables 1-3). The Cronbach's alpha coefficient for the first and the third factors for the high physician visit student subgroup was .69 and .71, respectively. Thus, overall the scale exhibited sufficient measurement reliability.
Discriminant Validity: WTCH Scale and Gender
Given some research evidence that suggests males and females differ in terms of health communication traits, with female patients talking more than male patients, self-disclosing more personal information about their health and their lifestyles within health settings (Gabbard-Alley, 1995), the WTCH scale should discriminate between men and women, such that there should be statistically significant differences between the two groups on the three factors of the scale. The results of three separate independent-sample t tests (see Table 4) showed significant differences on Factor 1 and Factor 3, with women indicating a higher willingness to communicate about health than men. The ability of the WTCH scale to discriminate between male and female tendencies toward willingness to communicate about health, thus, provides some support for the discriminant validity of the measure.
Convergent Validity: Relationship of WTCH Scale with Related Measures
Using the undergraduate student sample only to assess the convergent validity of the WTCH scale, each factor of the WTCH scale was correlated with each dimension of McCroskey and Richmond's (1987) Willingness to Communicate Scale (WTCS). In addition, to explore the relationships between the WTCH Scale, health information seeking, patient assertiveness, and patient nonadherence, the WTCH scale dimensions were correlated with Brashers et al.'s (1999) Patient Self-Advocacy Scale (PSAS).
Table 5 shows the correlations of each factor of the WTCH scale with each dimension of WTCS and PSAS. The results indicated that the willingness to communicate about health with providers dimension (Factor One) and the willingness to communicate with about health with nonproviders (Factor Two) of the WTCH scale were correlated with each dimension of WTCS, whereas Factor Three was correlated with only the willingness to communicate with strangers dimension of WTCS. The relationship between the WTCH scale and the PSAS was similar, with Factor One of the WTCH scale correlating with all three dimensions of PSAS, whereas Factor Two and Factor Three of the WTCH scale correlated only with the health information-seeking (education) dimension and the patient assertiveness dimension of the PSAS. Given the conceptual overlap between the dimensions of the WTCH, the dimensions of the PSAS, and the WTCS scale, the correlations between these dimensions provides some support for the convergent validity of the WTCH scale.
The purpose of this investigation was to develop and to test a measure of individuals' predisposition to communicate about health issues within the broader concept of the patient empowerment/self-advocacy paradigm of health. Toward that end, we developed the Willingness to Communicate about Health (WTCH) scale and administered it to two distinct samples. In this section, we discuss the implications of the study findings, identify some key limitations of the research and suggest directions for future research.
Conceptually, an individual's willingness to communicate about health with his or her health care provider should be related to his or her health information-seeking behaviors and level of assertiveness when communicating with that provider. Patient assertiveness and patient-centered communication are behaviors that are consistent with the larger idea of patient self-advocacy and empowerment (see Cegala, Post, & McClure, 2001; Vanderford, Jenks, & Sharf, 1997). If the WTCH Scale can be refined in future studies, it could serve as a tool for assessing the degree to which individuals may feel comfortable communicating about health issues with their providers, and it might ultimately be used to predict a patient's assertiveness and health information-seeking tendencies. Patients with low WTCH scores could be identified and perhaps educated about the potential benefits of becoming more assertive and proactive about obtaining health information during provider-patient interactions. Physicians would also potentially benefit from knowing such information, as they could use different communication strategies when interacting with low WTCH patients.
To make this instrument useful in these and others ways, it is important to examine the specific findings from this study and to refine the instrument based on those findings. As a starting point, although there were a number of differences in the findings of the factor analysis of the WTCH scale between the cancer patient sample and the undergraduate student sample, there were some important similarities. For example, for both samples, items 6, 8, 9, and 10 loaded on Factor One, and items 2 and 3 loaded on Factor Two. These similarities suggest that these particular items may be more stable in measuring these factors across different samples than the other items of the WTCH Scale. Factor Three was more problematic because it really only emerged as a factor for the student sample; however, item 1 was common to both samples for this particular factor.
There are number of possible explanations for the differences in findings across the two samples. First, the cancer patient sample was recruited from a cancer center where the participants had frequent interaction with physicians and other providers. In addition, each of these participants had been diagnosed (with a terminal cancer in most cases) and was currently seeking treatment for cancer during the time of study. It is very likely that the participants in the cancer sample had a much different view of their health situation than the undergraduate student sample, and this may have influenced their WTCH scores in ways that differed from the undergraduate student sample.
Moreover, the cancer patient sample consisted of only 95 participants, whereas the undergraduate student sample had 230 individuals, and this sample size difference may partially account for the different factor structures on the WTCH between the two samples. Finally, the cancer patient sample was much older than the undergraduate student sample, and this may have influenced the ways individuals conceived of the willingness to communicate about health construct and influenced ways in which participants in each sample responded to the scale items. For example, there is some evidence that suggests that older patients are sometimes more reluctant than younger patients to discuss health issues (Thompson, Robinson, & Beisecker, 2004).
Among the undergraduate student sample, dividing the sample into high and low physician visit groups indicated that the WTCH factors for students with a low number of physician visits were similar to the full sample of undergraduate students, whereas the WTCH factors were different for students with a high number of physician visits. This finding may indicate that individuals who see health care providers may more frequently interpret the WTCH scale items somewhat differently than those who see them less frequently. However, this needs to be assessed more thoroughly in future research.
The finding that the first two factors of the WTCH scale were correlated with each of the different dimensions of McCroskey and Richmond's (1987) Willingness to Communicate Scale, and that Factor Three of the WTCH was correlated with one of the dimensions of their scale, suggests that there may be some partial overlap between the two scales. Perhaps the relationship between the constructs of willingness to communicate about health and willingness to communicate in general have a relationship similar to other, more general trait measures vis-a-vis measures of the same trait in a more specific context, such as Wallston, Wallston, Kaplan, and Maides (1976) health locus of control scale that was developed from Rotter's (1954, 1975) general locus of control theory and scale. Wallston et al.'s (1976) scale provides more sensitive predictions of the relationship between internal health locus of control orientations and health behaviors than the more generalized locus of control scale. Similar to Wallston et al.'s (1976) scale, the WTCH may provide researchers with greater precision when assessing willingness to communicate within health contexts, in comparison to using the more general scale, although this needs to be assessed in future research.
The moderate positive correlation between Factor One of the WTCH and the health information-seeking dimension of the PSAS suggests that a person's willingness to communicate about his or her health with health care providers may be related to the amount of health information that person seeks and his or her level of assertiveness when communicating with those providers. The relationship between a person's willingness to communicate with providers and his or her tendency to not adhere to a provider's suggestions is less clear, particularly since there was a low and negative correlation between these two variables. Perhaps individuals who are more comfortable communicating with their physician ask more questions about suggested treatments, feel more comfortable about the suggested treatment option, and feel less need for noncompliance.
In addition, willingness to communicate with about health with nonproviders (Factor Two) and willingness to seek health information/discuss health issues (Factor Three) were positively correlated with health information-seeking behaviors and patient assertiveness. However, willingness to communicate about health with nonproviders (Factor Two) had relatively low correlations with these variables, whereas willingness to seek health information/discuss health issues (Factor Three) was moderately correlated with them. This suggests that a person's willingness to seek health information/discuss health issues may be related to his or her self-reported health information-seeking behaviors and level of assertiveness when communicating with providers. It would be useful in future research to assess whether WTCH scores on both dimensions are related to observed measures of health-information seeking and assertiveness behaviors (as opposed to self-reports).
Although this study was successful in constructing and testing the WTCH measure, a key limitation of the current study is that the WTCH scale for both the cancer patient and undergraduate student samples exhibited only relatively moderate reliability (ranging from .60 to .71; average reliability across factors for both samples equals .67). This finding suggests that there may have been some confusion among the participants regarding how to interpret the WTCH Scale items. Because the cancer patients in this study were recruited from a cancer center, some of these individuals may have interpreted the scale items within the context of the cancer center, whereas others interpreted the items to apply to their willingness to communicate about health outside of the center. Among the undergraduate students, there may have been interpretation problems as well, and it is difficult to know how seriously or accurately they responded to the items.
The current study represents a first step towards creating an instrument to measure people's tendency to communicate about health with providers, nonproviders, and when information about health concerns is needed. Future research needs to assess whether the WTCH scale yields similar findings to the current study across different types of populations. We will seek in future work to refine the measure and ultimately improve the reliability and validity of the measure. Such an effort may help providers identify at risk patients, and interventions could be developed to help them to be more assertive during medical encounters.
Patient self-advocacy is an important concept in the larger framework of patient-center communication and patient empowerment. Patient-self advocacy rests on the assumption that individuals are willing to communicate about their health concerns with providers and with other people within their social networks. The WTCH Scale attempts to measure an individual's tendency to discuss health issues with health care providers and members of his or her social network. We hope that measuring this important communication trait will help health communication researchers and practitioners alike to identify people who are reluctant to discuss health issues, so that interventions can be created to help empower these individuals and to educate them about the significant benefits that accrue from engaging in greater self-advocacy when seeking health care.
This manuscript was accepted by the previous editor, Professor Jim L. Query.
Avtgis, T. A. (1999). The relationship between unwillingness to communicate and family communication patterns. Communication Research Reports, 16, 333-338.
Balint, J. & Shelton, W. (1996). Regaining the initiative: Forging a new model of the patient-physician relationship. Journal of the American Medical Association, 275, 887-892.
Ballard-Reisch, D. S. (1990). A model of participative decision making for physician-patient interaction. Health Communication, 2, 91-104.
Beatty, M. J. (1998). Future directions in communication trait theory and research. In J. C. McCroskey, J. A. Daly, M. M. Martin, & M. J. Beatty (Eds.), Communication and personality: Trait perspectives (pp. 41-67). Cresskill, NJ: Hampton Press.
Beisecker, A. E. (1990). Patient power in doctor-patient communication: What do we know? Health Communication, 2, 105-122.
Brashers, D. E., Haas, S. M., & Neidig, J. L. (1999). The patient self-advocacy scale: Measuring patient involvement in health care decision-making interactions. Health Communication, 11, 97-121.
Brashers, D. E. & Klingle, R. S. (1992, October). The influence of activism on physician-patient communication. Paper presented at the meeting of the Speech Communication Association, Chicago, IL.
Brown, J. B., Stewart, M., & Ryan, B. L. (2003). Outcomes of patient-provider interaction. In T. L. Thompson, A. M. Dorsey, K. I. Miller, & R. Parrot (Eds.), Handbook of health communication (pp. 141-162). Mahwah, NJ: Lawrence Erlbaum.
Burgoon, J. K. (1976). The unwillingness-to-communicate scale: Development and validation. Communication Monographs, 43, 60-69.
Cegala, D. J., Post, D., & McClure, L. (2001). The effects of patient communication skills training on the discourse of elderly patients during a primary care interview. Journal of the American Geriatrics Society, 49, 1505-1511.
Crowell, T. L. (2004). Seropositive individuals' willingness to communicate, self-efficacy, and assertiveness prior to HIV infection. Journal of Health Communication, 9, 345-424.
Daly, J. A. & Diesel, C. A. (1992). Measures of communication-related personality variables. Communication Education, 41, 405-414.
DiMatteo, M. & Hays, R. (1981). Social support and serious illness. In B. Gottlieb (Ed.), Social networks and social support (pp. 117-148). Beverly Hills, CA: Sage.
Gabbard-Alley, A. S. (1995). Health communication and gender: A review and critique. Health Communication, 7, 35-54.
Golin, C. E., DiMatteo, M. R., & Gelberg, L. (1996). The role of patient participation in the doctor visit: Implications for adherence to diabetes care. Diabetes Care, 19, 1153-1164.
Gotcher, J. M. & Edwards, R. (1990). Coping strategies of cancer patients: Actual communication and imagined interactions. Health Communication, 2, 255-266.
Greenfield, S., Kaplan, S. H., & Ware, J. E., Jr. (1985). Expanding patient involvement in care: Effects on patient outcomes. Annals of Internal Medicine, 102, 520-528.
Infante, D. A., Rancer, A. S., & Womak, D. F. (2003). Building communication theory (4th ed.). Prospect Heights, IL: Waveland Press.
Keyton, J. & Frey, L. R. (2002). The state of traits: Predispositions and group communication. In L. R. Frey (Ed.), New directions in group communication (pp. 99-120). Thousand Oaks, CA: Sage.
Levinson, W., Gorawara-Bhat, R., Dueck, R., Egener, B., Kao, A., Kerr, C., et al. (1999). Resolving disagreements in the patient-physician relationship: Tools for improving communication in managed care. Journal of the American Medical Association, 282, 1477-1483.
Lin, Y. & Rancer, A. S. (2003). Ethnocentrism, intercultural communication apprehension, intercultural willingness-to-communication, and intentions to participate in an intercultural dialogue program: Testing a proposed model. Communication Research Reports, 20, 62-72.
McCroskey, J. C. (1992). Reliability and validity of the willingness to communicate scale. Communication Quarterly, 40, 16-25.
McCroskey, J. C. & Richmond, V. P. (1982). Communication apprehension and shyness: Conceptual and operational distinctions. Central States Speech Journal, 33, 458-468.
McCroskey, J. C. & Richmond, V. P. (1985, March). Willingness to communicate and interpersonal communication. Paper presented at the West Virginia Symposium on Personality and Interpersonal Communication, Morgantown, WV.
McCroskey, J. C. & Richmond, V. P. (1987). Willingness to communicate, in J. C. McCroskey & J. A. Daly (Eds.), Personality and interpersonal communication (pp. 189-156). Newbury Park, CA: Sage.
McCroskey, J. C. & Richmond, V. P. (1990). Willingness to communicate: Differing cultural perspectives. Southern Communication Journal, 56, 72-77.
McCroskey, J. C. & Richmond, V. P. (1998). Willingness to communicate. In J. C. McCroskey, J. A. Daly, M. M. Martin, & M. J. Beatty (Eds.), Communication and personality: Trait perspectives (pp. 118-131). Cresskill, NJ: Hampton Press.
Menzel, K. E. & Carrell, L. J. (1999). The impact of gender and immediacy on willingness to talk and perceived learning. Communication Education, 48, 31-40.
Morgan, S. E. & Miller, J. K. (2002). Beyond the organ donor card: The effect of knowledge, attitudes, and values on willingness to communicate about organ donation to family members. Health Communication, 14, 121-134.
Mortensen, D. C., Arnston, P. H., & Lustig, M. (1977). The measurement of verbal predispositions: Scale development and application. Human Communication Research, 3, 146-158.
Richmond, V. P. & Roach, K. D. (1992). Willingness to communicate and employee success in U.S. organizations. Journal of Applied Communication Research, 20, 95-115.
Rocca, K. A. & Martin, M. M. (1998). The relationship between willingness to communicate and solidarity with frequency, breadth, and depth of communication in the sibling relationship. Communication Research Reports, 15, 82-90.
Rotter, J. B. (1954). Social learning and clinical psychology. Englewood Cliffs, NJ: Prentice-Hall.
Rotter, J. B. (1975). Some problems and misconceptions related to the construct of internal versus external control of reinforcement. Journal of Consulting and Clinical Psychology, 43, 56-67.
Smith, S. W., Kopfman, J. E., Massi-Lindsey, L. L., Yoo, J., & Morrison, K. (2004). Encouraging family discussion on the decision to donate organs: The role of the willingness to communicate scale. Health Communication, 16, 333-346.
Thompson, T. L., Robinson, J. D., & Beisecker, A. E. (2004). The older patient-physician interaction. In J. F. Nussbaum & J. Coupland (Eds.), Handbook of communication and aging research (2nd ed., pp. 451-477). Mahwah, NJ: Lawrence Erlbaum.
Vanderford, M. L., Jenks, E. B., & Sharf, B. F. (1997). Exploring patients' experiences as a primary source of meaning. Health Communication, 9, 13-26.
Wallston, B. S., Wallston, K. A., Kaplan, G. D., & Maides, S. A. (1976). Development and validation of the health locus of control (HLC) scale, Journal of Consulting and Clinical Psychology, 44, 580-585.
Wright, K. B. & Frey, L. R. (in press). Cancer support groups. In H. D. O'Hair, L. Sparks, & G. L. Kreps (Eds.), Health communication and cancer. Cresskill, NJ: Hampton Press.
Young, A. & Flower, L. (2002). Patients as partners, patients as problem solvers. Health Communication, 14, 69-97.
Correspondence to: Kevin B. Wright, Ph.D., Department of Communication, 610 Elm Avenue, Room 101, University of Oklahoma, Norman, OK 73019.
Table 1 Cancer Patient Sample: Standardized Factor Loadings, Reliability (Cronbach's [alpha]), and Correlation Among Factors for Willingness to Communicate About Health Scale for Full Sample (N = 95) All items Factor name and scale item I II III 1: Factor 1 [alpha] = .71 Item 8: Comfortable talking about .78 .09 .00 health with health care providers Item 5: Actively seek out .66 .04 -.27 information about health Item 6: Quick to make an .51 .03 -.18 appointment to talk with physician when not well Item 9: Experience difficulties .48 -.19 .00 communicating successfully with health care providers Item 10: Competent communicator .47 -.14 .17 when talking 'about health issues 11: Factor 2 [alpha] = .69 Item 3: Comfortable talking about .05 -.84 -.07 their health with a variety of people not counting physicians Item 4: Only talk about health .02 -.53 -.05 issues when have to Item 2: Comfortable talking about .19 -.47 -.39 my health with a wide variety of people not counting physicians Item 7: When don't feel well, don't -.03 -.35 .01 want to talk III: Factor 3 Item 1: 1 Frequently talk about .01 -.22 -.58 health issues Correlation (r) among factors I-II: -.42 I-III: -.21 I1-III:.19 Items eliminated Factor name and scale item I II 1: Factor 1 [alpha] = .71 Item 8: Comfortable talking about .82 -.15 health with health care providers Item 5: Actively seek out .68 .01 information about health Item 6: Quick to make an .54 -.02 appointment to talk with physician when not well Item 9: Experience difficulties .43 .18 communicating successfully with health care providers Item 10: Competent communicator .43 .15 when talking 'about health issues 11: Factor 2 [alpha] = .60 Item 3: Comfortable talking about .16 .67 their health with a variety of people not counting physicians Item 4: Only talk about health .03 .59 issues when have to Item 2: Comfortable talking about -- -- my health with a wide variety of people not counting physicians Item 7: When don't feel well, don't -.05 .42 want to talk III: Factor 3 Item 1: 1 Frequently talk about -- -- health issues Correlation (r) among factors I-II:.39 Table 2 Student Sample: Standardized Factor Loadings, Reliability (Cronbach's [alpha]), and Correlation Among Factors for Willingness to Communicate About Health Scale for Full Sample (N = 230) Full sample Factor name and scale item I II III I: Factor 1 [alpha] = .66 Item 8: Comfortable talking about health .66 .03 -.03 with health care providers Item 9: Experience difficulties .59 .03 .13 communicating successfully with health care providers Item 6: Quick to make an appointment to .50 -.23 -.29 talk with physician when not well Item 10: Competent communicator when .48 .21 -.03 talking about health issues Item 7: When don't feel well, don't want .35 .05 -.09 to talk II: Factor 2 [alpha] = .60 Item 2: Comfortable talking about my .10 .67 -.10 health with a wide variety of people not counting physicians Item 3: Comfortable talking about their .03 .54 -.04 health with a variety of people not counting physicians III: Factor 3 [alpha] = .67 Item 5: Actively seek out information .02 -.09 -.83 about health Item 1: 1 Frequently talk about health -.02 .17 -.58 issues Item 4: Only talk about health issues .07 .19 -.38 when have to Correlation (r) Between Factors I-II: .32 I-III -.47 II-III: -.30 Table 3 Student Sample: Standardized Factor Loadings, Reliability (Cronbach's [alpha]), and Correlation Among Factors for Willingness to Communicate About Health Scale for Low (n = 158) and High (n = 72) Physician Visits Subgroups Low physician visits Factor name and scale item I II III I: Factor 1 [alpha] = .65 Item 8: Comfortable talking about health .62 .01 -.05 with health care providers Item 9: Experience difficulties .67 .15 -.08 communicating successfully with health care providers Item 6: Quick to make an appointment to .45 -.32 .16 talk with physician when not well Item 10: Competent communicator when .45 -.01 -.32 talking about health issues Item 7: When don't feel well, don't want .38 -.09 .06 to talk II: Factor 2 [alpha] = .65 Item 2: Comfortable talking about my .13 -.64 -.11 health with a wide variety of people not counting physicians Item 3: Comfortable talking about their -.06 -.69 -.04 health with a variety of people not counting physicians III: Factor 3 [alpha] = .66 Item 5: Actively seek out information -.06 .00 -.85 about health Item 1: 1 Frequently talk about health -.02 -.07 -.62 issues Item 4: Only talk about health issues .13 -.18 -.34 when have to Correlation (r) Between Factors I-II: .31 I-III: -.39 II-III: -.23 High physician visits Factor name and scale item I II III I: Factor 1 [alpha] = .64 Item 8: Comfortable talking about health .57 .03 -.23 with health care providers Item 9: Experience difficulties .44 .01 .06 communicating successfully with health care providers Item 6: Quick to make an appointment to .66 .08 .11 talk with physician when not well Item 10: Competent communicator when .53 -.12 -.06 talking about health issues Item 7: When don't feel well, don't want .35 -.24 -.16 to talk II: Factor 2 [alpha] = .55 Item 2: Comfortable talking about my .01 -.19 -.56 health with a wide variety of people not counting physicians Item 3: Comfortable talking about their -.04 -.94 -.02 health with a variety of people not counting physicians III: Factor 3 [alpha] = .64 Item 5: Actively seek out information .43 .05 -.32 about health Item 1: 1 Frequently talk about health .11 -.22 -.65 issues Item 4: Only talk about health issues -.03 .13 -.71 when have to Correlation (r) Between Factors Table 4 Student Sample: Mean, (Standard Deviation), Sample Size, and t Statistic for Willingness to Communicate about Health Scale (WTCHS) by Gender WTCHS factors Male (n = 92) Female (n = 139) t (df) M (SD) M (SD) Factor 1 3.46 (.53) 3.72 (.60) 3.39 * (229) Factor 2 3.57 (.73) 3.60 (.79) .29 (229) Factor 3 3.09 (.77) 3.43 (.79) 3.24 * (229) * P < .01. Table 5 Student Sample: Correlations among Willingness to Communicate About Health Scale (WTCHS) Factors and Dimensions of Patient Self-Advocacy Scale (PSAS) and Willingness to Communicate Scale (WTCS) WTCHS PSAS factors Education Assertiveness Nonadherence Factor 1 .43 ** .43 ** -.22** Factor 2 .23 ** .14 * -0.04 Factor 3 .49 ** .40 ** .03 WTCHS WTCS factors Interpersonal Stranger Acquaintance Friend Factor 1 .18 ** .21 ** .20 ** .18 ** Factor 2 .19 ** .18 ** .15 * .14 * Factor 3 .09 .17 * .05 .04 N = 230; * p < .05; ** p < .01 level (two-tailed).…