Clinical, classroom, or personal education: attitudes about health literacy *

to address: (a) the relative beneﬁts of launching health literacy initiatives in clinical settings, classroom, or personal educational venues and (b) the ways the perceptions of health literacy are impacted by a health professional’s characterization of the people they serve as ‘‘patients,’’ ‘‘students,’’ or ‘‘consum-ers.’’

Findings: Three factors were identified. Factor 1 is optimistic and supportive of health literacy's transformative sociocultural and professional potential, if clinical settings become a launching point for health literacy activities. Factor 2 is less optimistic about health literacy's potential to improve clinical or patient outcomes and prefers to focus health literacy initiatives on classroom education settings. Factor 3 supports improving the nation's health literacy but tends to support health literacy initiatives when people privately interact with health information materials.
Conclusions: Each factor's attitudes about the appropriate educational venue to initiate health literacy activities are different and somewhat mutually exclusive. This suggests that health literacy is seen through different perceptual frameworks that represent a possible source of professional disagreement.

INTRODUCTION The Institute of Medicine's report on health literacy: its impact and critique
In its recent report, Health Literacy: A Prescription to End Confusion, the Institute of Medicine (IOM) defined health literacy as ''the degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions'' [1]. The IOM reported that even well-educated Americans have problems understanding medical jargon, medical service forms, and prescription information as well as following navigational directions in some hospitals and medical centers. Marcus characterized the nation's poor health literacy as public health's ''silent epidemic'' [2].
After reviewing anecdotal and statistical evidence, the IOM concluded that improving health literacy is one of the most pressing health care delivery and health policy challenges facing the US health care delivery system [1]. The former US surgeon general noted that improving the nation's health literacy should be a national health and public policy priority [3].
Numerous researchers agree that health literacy (as defined by the IOM) represents an important, even foundational, idea that weaves together a significant array of issues in health policy, health services research, health communication, and health care delivery [4][5][6][7][8][9][10]. Issues in health policy, health care delivery,

Figure 1
Issues related to health literacy Ⅲ Reducing health disparities or improving the access and delivery of care to African Americans, Hispanic Americans, Native Americans, and other medically underserved groups Ⅲ Improving adult understanding of quantitative and visual medical information Ⅲ Tailoring health care materials to audience literacy levels Ⅲ Communicating health information in jargon-free, easy-to-understand language Ⅲ Enhancing patient empowerment and compliance with medical instructions Ⅲ Providing more patient-centered health care delivery Ⅲ Improving patient-provider communication Ⅲ Focusing on disease prevention instead of treatment to reduce health care costs Ⅲ Providing better health education for adults and for K-12 students Ⅲ Improving patient compliance Ⅲ Performing better patient screening to assess health literacy capabilities Ⅲ Committing fewer errors in hospitals and health care institutions Ⅲ Improving health services utilization Ⅲ Improving health care outcomes for patients [5,6,8,[11][12][13][14][15] and health services that have become associated with health literacy include reducing health disparities and clarifying and enhancing the presentation of health information ( Figure 1) [5,6,8,[11][12][13][14][15].
However, as soon as the richness of health policy issues associated with health literacy were identified and improving health literacy was embraced as a priority, some uneasiness about the pace of and readiness to initiate health literacy activities surfaced in the academic literature and some mailing lists for health professionals. Although there was little disagreement about (a) the data that described the nation's health literacy in the IOM report, (b) the idea that Americans should be able to understand medical instructions, and (c) the assertion that consumers should receive optimal access to evidence-based health information, some critics in 2004 to 2005 questioned if it was premature for health care professionals, medical associations, and governmental and nongovernmental organizations to make significant investments in health literacy initiatives.
For example, Tones [16] noted the IOM's definition of health literacy might be clear, but its rhetorical interpretation and implementation were so far-reaching that the term was losing its clarity and purpose. Bass [5] found it was difficult to pinpoint the precise meaning of the term ''health literacy,'' and Speros [17] implied it was a challenge to operationalize the term into research variables.
McCray [6] and Lee [13] added that it might be premature for government, medical associations, or others to advance health literacy as a significant health policy priority because its evidence base is in a formative (as opposed to an advanced) stage. McCray [6] found research needs to be developed in several areas, including the interactions among general literacy, health literacy, information technologies, and the existing health care infrastructure.
In addition to scholarly reservations, a range of informal criticisms about health literacy surfaced on mailing lists populated by health care, medical, and public health professionals in 2005. Mailing lists or online forums are increasingly important to researchers as venues to track the range of public and especially professional discourse regarding topics of current concerns [18].
Many of the mailing list postings discussed what the intention of health literacy initiatives was or what health literacy research and public intervention campaigns might strive to accomplish. Other postings proposed definitions of health literacy, limiting or expanding the term's conceptual range and the wisdom of a broadly or narrowly based definition. Some of these postings speculated on the long-range health policy, patient outcomes, and sociocultural impacts of health literacy initiatives. Some of the more sociologically grounded discussion questioned, for example, if assigning people into marginal and low literacy categories might undermine the cultural acceptability of health literacy campaigns in both groups.
By spring 2005, the aggregate of the mailing list and scholarly criticisms suggested health literacy was greeted with support as well as some uncertainty and apprehension by some health professionals. More recently, Parker and Kindig [19] asked if the IOM's health literacy recommendations were being taken seriously by medical professionals.
Overall, the literature and mailing list comments suggest that many issues surrounding health literacy are salient to health care professionals, but opinions differ about the pace and appropriate strategies to advance health literacy initiatives. These differences could impede its momentum.

Problem statement
In turn, a current question among health literacy advocates and critics is: How do health care professionals assess some of the health literacy opinions expressed by their peers? Do contentious or supportive opinions about health literacy resonate with health care professionals engaged in thinking about health literacy?
This study seeks to provide insights into how people with a professional reason to be conscientious about health literacy assess a spectrum of favorable and unfavorable attitudes about some of the ideas, definitions, strategies, tactics, and long-range social implications that the IOM report raised. More specifically, this study seeks to better understand how a spectrum of attitudes about health literacy are perceived by medical librarians and environmental health experts from public health programs and medical colleges that historically assist African Americans, Hispanic Americans, and Native Americans.
The study examines how opinion clusters are forming among two of the professional groups engaged in an evolving debate about an important health policy issue. The broader judgments and attitudes of these groups may be critical to health literacy's evolution and development. The array of opinions found in the study attempts to reflect the operant dialogue and subjectivity in summer 2005 (the time this research was initiated) among some interested professionals.
The study does not explore a specific hypothesis about what health literacy research and campaign initiatives should strive to accomplish or a preferred con-ceptual definition of health literacy. The study does not evaluate how respondents project the impact of health literacy initiatives on patients or society. Instead, using Q methodology, the study explores how its respondents assess diverse opinions taken from a professional discourse regarding health literacy.

Explanation of Q methodology
A mixed method employed in numerous disciplines [20][21][22][23][24][25], Q methodology uses quantitative methods to determine factor arrays and qualitative judgment to interpret a factor's broader meaning [20][21][22][23]. Brown [20] explains that Q methodology is ideally used to explore how diverse opinions are structured. In contrast to most public opinion research, respondents in Q studies ideally are well informed and directly involved in the area under study. The issues raised in the survey instrument are also highly salient [20,21,[23][24][25].
Brown [20] and Smith [25] explain that other important differences between social science opinion measurement approaches and Q methodology include the latter's procedural embellishment that asks participants to sort through their opinions in a balanced, quasi-normal distribution after scoring each item individually. As a result, each factor actually represents how some of the participants sorted all the statements in common.
Another major difference is that factor analysis is used in Q methodology as an exploratory technique to discover how opinions are clustered or segmented in groups (factors). In traditional social science opinion measurement, factor analysis is often used to confirm if factor patterns are consistent with predetermined hypotheses or expected patterns. In Q methodology, an investigator avoids prejudgments by deliberately not positing hypotheses or research questions.
Researchers also are encouraged to explain how factor scores articulate a distinguishing thematic voice, irrespective of whether they fit in the predetermined categories or expectations that fostered the survey instrument. Hence, factor descriptions emphasize each factor's distinguishing perceptual perspective or discerning characteristics.
Finally, in Q methodology the ''n'' is not the number of respondents, but the number of respondents multiplied by the number of items in the survey that were sorted. In the current study, the number of responses is n ϭ 1,836 (51 respondents multiplied by 36 statements).

Sample
An online survey of thirty-six items using Q methodology with fifty-one nonrandomly chosen respondents was conducted in two phases in spring 2005 and winter 2006. Respondents self-selected to participate and were invited from a nonrandom, convenience sample of two groups. One invited group included members of the Consumer and Patient Health Information Section of the Medical Library Association (MLA). Given the group's interest in health literacy issues, all members of the Consumer and Patient Health Information Section of MLA were invited to participate through an email sent by the president and president-elect of MLA and MLA's executive director in early February 2006.
The second invited group consisted of members of the Environmental Health Information Outreach Program at the US National Library of Medicine. This is an advisory group of environmental health experts representing schools of medicine and public health that specialized in assisting medically underserved audiences. At the time of the study, the topic of health literacy was current among the public health programs and medical colleges that the group represented [15,26,27]. Participation was voluntary; no incentives were provided. The study was exempt from review by the US Office of Management and Budget.

Data collection
Similar to the process of assembling the items or statements for a Q sort as described by Brown [20], the instrument used in this study was derived primarily from verbatim statements of opinion about health literacy. In this case, statements of opinion (called ''statements'' throughout the rest of the manuscript) were derived from the literature and mailing lists. Verbatim statements were sampled from mailing lists where comments from health care professionals about the IOM report and health literacyrelated issues had been expressed. Two mailing lists were monitored: Literacy NIFL-HEALTH Forum Ͻ h t t p : / / w w w. n i fl . g ov / l i n c s / d i s c u s s i o n s / nifl-health/healthliteracy.htmlϾ and a Floridabased mailing list Ͻhttp://www.floridaliteracy.org/ discussionlinks.htmlϾ. Most of the statements used in the study were sampled from the NIFL-HEALTH Forum mailing list. Some statements about health literacy were sampled verbatim from the literature to obtain a more diverse spectrum of opinion.
While some statements from the published literature were taken from articles written before the release of the IOM report toward the end of 2003, all comments from mailing lists were taken from postings in 2004 and 2005. The original corpus of almost 500 statements about health literacy was reduced to 36 statements.
Following a process recommended by McKeown and Thomas [21] and Stephenson [28], statements were organized into three broad discourse themes: 1. What should be the primary intention of future health literacy initiatives? What should health literacy research and campaign initiatives primarily strive to accomplish? 2. What is health literacy? How should health literacy be conceptually defined? 3. In the long run, what will be the primary health policy, cultural, patient outcomes, and impacts of health literacy initiatives?
Of the thirty-six statements selected for the final instrument, twelve represented each of the three themes and were balanced to reflect opposing points of view as recommended by McKeown and Thomas [21]. Opposing points of view were operationally defined as assertions favorable or unfavorable to the IOM report's suggestions and its projected social and clinical outcomes.
Although the author used verbatim statements expressed in natural language whenever possible, some statements were edited for clarity after a pretest of the proposed instrument. A pretest of the instrument with eighteen officials from several US federal health agencies, who have a professional interest in health literacy, occurred in summer 2005.
Respondents were given a three-week period to complete the instrument online. The members of the Environmental Health Information Outreach Program completed the survey from mid-June to mid-July 2005. MLA members completed the instrument from February to March 2006.
To administer the instrument online, WebQ, a program that permits Q sorts to be placed and completed on the Internet, was modified. The instrument was hosted by a server with password access at the Lister Hill National Center for Biomedical Communications, US National Library of Medicine, National Institutes of Health. PCQ, a commercial software package designed for Q methodology, was used for data entry and analysis.
As recommended by McKeown and Thomas [21], respondents assessed each of the 36 statements in the instrument in 2 separate steps. First, respondents assessed each statement individually in a Likert scale from Ϫ4 to ϩ4 (representing a spectrum from strongly disagree to strongly agree). Second, respondents ranked, or intercompared, their opinions of all 36 statements on the same scale. The factor scores, reported in Table 1, are based on the latter sort.

Data analysis
Respondents' sorting of all 36 statements were correlated, and 3 factors were derived from a principal components factor matrix, subject to a varimax rotation. The number of factors was determined by an Eigenvalue of greater than 1.0. The Guilford-Lacey expression was used to determine significant factor loadings, in this case greater than 0.40, which is statistically significant (P Ͻ 0.05). Weightings based on each respondent's factor loadings were applied to rankings of individual statement so that factors could be represented as normalized arrays, or Z scores.
Three factors, or patterns of opinion segmentation, were interpreted from the normalized factor arrays, which are reported in Table 1. A key to interpreting Table 1 is provided in Figure 2.
Factors were interpreted by identifying persistent themes and patterns as well as the items or statements in the instrument that distinguished each factor's perspective, as recommended by McKeown and Thomas [21]. The label assigned to each factor summarizes each factor's most distinguishing judgmental pattern or point of view.

RESULTS
Results are presented selectively for each factor.

Respondent demographics
Fifty-one completed surveys were received. Incomplete surveys were discarded. Demographic information is summarized in Table 2 and presented in full in Table 3.

Survey development
Points of view reflected in the survey statements were coded as favorable or unfavorable to the IOM's health literacy report's finding. The intercoder reliability, determined by a formula suggested by Holsti [29], for coding statements as critical or favorable was 92% for 2 coders.
In a post-Q sort question, all fifty-one respondents either agreed or strongly agreed with the statement, ''I am personally interested in health literacy.'' Fortysix of the fifty-one respondents agreed or strongly agreed with the statement, ''I read extensively about health literacy.'' These results suggested face validity to the assumption that health literacy issues were salient to respondents.

Factor constituents
All three factors accounted for 44% of the study's possible variance. Factor 1 accounted for 17% of the variance, factor 2 for 12%, and factor 3 for 15% of the variance.
Of the fifty-one respondents, fourteen people loaded significantly on factor 1, eight loaded significantly on factor 2, and twelve loaded significantly on factor 3. Twelve respondents were ''confounded,'' or had multiple loadings on more than one factor. Q sorts from five respondents were not significant (Table 3). Table 3 contains summary information about each factor's constituents. Because the emphasis in Q methodology is on psychographic archetypes rather than demographic differences [20], the focus of the discussion in the following sections is on each factor's distinguishing thematic pattern.

Factor descriptions
Factor 1: clinical and patient orientation. Table 1 lists all statements and factor loadings. Factor 1 is partially distinguished by its endorsement of health literacy initiatives that are designed to boost patient cognitive skills. For example, factor 1 strongly agrees or agrees with these statements: Ⅲ The primary intent of health literacy initiatives is to empower patients Ⅲ The primary intent of health literacy initiatives should be to improve the ability of consumers/patients/caregivers to think critically about the health information they receive Ⅲ The primary goal for health literacy initiatives should be to increase the use of plain language in all consumer instructions and communication about health 1. The primary intent of health literacy initiatives should be to improve the ability of consumers/patients/ caregivers to think critically about the health information they receive.

2
Ϫ4 4 2. The foundation of health literacy is that all communications between patients/caregivers/families/ consumers and providers need to be conceived as a dialogue.

1
Ϫ1 2 3. The primary intent of health literacy initiatives is to help patients better understand what a physician tells them as well as prescription information.

4
Ϫ3 1 4. Health literacy is a concept that fails to capture the public's attention or imagination. 1 0 Ϫ2 5. Health care costs will decline as a result of improving health literacy.
3 0 Ϫ2 6. Health literacy efforts are a preemptive strike designed to reduce the liabilities of physicians, hospitals, and insurance carriers when patients do not understand the information presented to them.
Ϫ4 1 Ϫ4 7. Poor K-12 health education programs are the major barrier to improving health literacy. Ϫ1 3 Ϫ1 8. Teaching health literacy to help consumers negotiate the health care system is a superficial fix (or ''too little too late'') because it fails to address the larger problems underlying the US health care system (such as 45 million uninsured Americans). Ϫ4 Ϫ2 Ϫ2 9. The concept of health literacy has become overburdened with too many sociological, cultural, psychological, health communication, and demography concepts.
10. Writing health materials at a sixth-grade level dilutes the quality of needed health information for all consumers.
11. It is naive to use public health or disease intervention conceptual models to resolve cultural challenges, such as health literacy.
Ϫ3 0 Ϫ3 12. Health literacy educational results will not be immediate. A generation or more may be needed to effect cultural changes for both caregivers and consumers.    Ϫ4 3 27. Since health literacy efforts have not been generated at a grassroots level, they appear to be culturally paternalistic. Ϫ3 Ϫ1 Ϫ2 28. Health literacy should conceptually shift its focus from improving comprehension skills to urging disease prevention and early detection. 29. Health literacy initiatives should help more people of all ages understand how the human body functions rather than focusing on lay translations of medical terms and vocabulary.
Ϫ2 1 0 30. Health literacy initiatives should focus on helping patients, caregivers, and consumers better navigate the health care delivery system.

2
Ϫ1 3 31. The concept of health literacy needs to better encompass the stress (even the panic) people feel when they need to know and then are left to seek health information.
2 Ϫ3 2 32. Health literacy efforts will help amend the shame and stigma associated with limited literacy skills in American society because health literacy efforts provide a comprehensive, sustained effort to reduce these differences.  Guide to reading the table: Read down a column to interpret each factor's set of responses for the entire instrument. Read across the row to interpret how each factor answered each statement differently. The scores provided are each factor's Z score (or a composite response) for each statement. A description of how each Z score was interpreted is provided in the manuscript's methods section. Ⅲ Health literacy initiatives should focus on helping patients, caregivers, and consumers better navigate the health care delivery system Similarly, factor 1 perceives that improving patient cognitive skills has pragmatic economic and psycho-logical advantages. Factor 1 agrees or slightly agrees that: Ⅲ Health literacy should conceptually encompass how patient misunderstandings and confusion add hidden costs to the nation's health care delivery system Interpretation of Z scores in Table 1 Ⅲ Scores of Ϫ4 are qualitatively interpreted as representing a strong disagreement with a statement Ⅲ Scores of Ϫ3 and Ϫ2 are interpreted as representing disagreement Ⅲ Scores of Ϫ1 are interpreted as representing a slight disagreement. Ⅲ Z Scores of ϩ4 are qualitatively interpreted as representing strong agreement Ⅲ Scores of ϩ3 and ϩ2 are interpreted as representing agreement Ⅲ Scores of ϩ1 are interpreted as representing a slight agreement Ⅲ Scores of 0 are interpreted as representing a neutral view (neither agreeing or disagreeing) MLA ϭ Medical Library Association. EnHIOP ϭ Environmental Health Information Outreach Program. Ⅲ The concept of health literacy needs to better encompass the stress (even the panic) people feel when they need to know and then are left to seek health information Factor 1's broad support for patient empowerment and cognitive development is accompanied by an interest in improving patient interaction with physicians and other health care providers. Factor 1 strongly agrees, agrees, or slightly agrees that: Ⅲ The primary intent of health literacy initiatives is to help patients better understand what a physician tells them as well as prescription information Ⅲ The primary intent of health literacy initiatives is to improve patient adherence with physicians' instructions Ⅲ Improving health literacy will first result in patients who are more discerning about the medical advice they receive Also, factor 1's interest in empowering patients (to enhance provider-patient interactions) has some limitations. Factor 1 is uncomfortable with a clinical environment where the perceived locus of control shifts from providers to patients. For example, factor 1 slightly agrees or is neutral about these statements: Ⅲ The concept of health literacy should be expanded to include an emphasis on consumer skills to access health services and to engage in patient advocacy Ⅲ The foundation of health literacy is that all communications between patients/caregivers/families/ consumers and providers need to be conceived as a dialogue Besides supporting clinical settings to advance health literacy, factor 1 is less supportive of some nonclinical venues, which are endorsed by the other two factors. For example, factor 1 does not believe that broad health education programs or efforts to improve K-12 health instruction are a panacea to improve the nation's health literacy. Factor 1 disagrees or slightly disagrees that: Ⅲ The primary intention of health literacy initiatives should be to help primary or secondary school educators teach health literacy as a basic subject (e.g., such as English or math) Ⅲ Poor K-12 health education programs are the major barrier to improving health literacy Ⅲ Health literacy initiatives should help more people of all ages understand how the human body functions rather than focusing on lay translations of medical terms and vocabulary This set of responses illustrates a conceptual dis-tinction between factor 1's and factor 2's attitudes about the appropriate venue to address health literacy challenges and launch initiatives. Factor 1 seems more interested in health literacy initiatives that support patients in a clinical environment than health literacy efforts aimed at students in a classroom setting. Unlike factor 2, factor 1 also is skeptical that a conceptual framework derived from public health should underlie how health literacy initiatives are conceived. Factor 1 either disagrees or slightly disagrees that: Ⅲ Health literacy should conceptually shift its focus from improving comprehension skills to urging disease prevention and early detection Ⅲ The primary intent of health literacy initiatives should better document the causal pathway of how poor literacy affects health In addition, despite their support for improving patient cognition, factor 1 does not endorse some health literacy initiatives designed to improve a consumer's personal general medical knowledge or education. Factor 1's lack of enthusiasm regarding questions related to improving a consumer's personal education differentiates their attitudes from those endorsed by factor 3. For example, factor 1 slightly disagrees or is neutral about the following statements: Ⅲ Foremost, health literacy should be conceptually conceived as improving a consumer's basic medical knowledge Ⅲ The concept of health literacy should focus on a consumer's skill to interpret media messages, enable people to look for (and assess whom to ask for) more health information Ⅲ The primary intent of health literacy initiatives should be to provide demographically targeted, justin-time medical information Overall, factor 1's emphasis on advancing health literacy initiatives seems directed at improving primary care delivery for patients during clinical interactions. Factor 1 is also less enthusiastic about alternative delivery points, such as through school instruction or via personal-educational targeted media. To factor 1, apparently the ''teachable moment'' where health literacy initiatives become viable is when adults encounter clinical care and the health care delivery system.
Although it is not one of the perceptual focal points that distinguishes factor 1 from factor 2 or factor 3, it should be added that factor 1 is optimistic about the public impact of health literacy initiatives. For example, factor 1 strongly agrees or agrees that: Ⅲ Elevating the nation's health literacy is a vital step to improving the quality of health care and health outcomes Ⅲ Health care costs will decline as a result of improving health literacy Similarly, factor 1 rejects forecasts that health literacy initiatives might have some deleterious sociocultural consequences. For example, factor 1 strongly disagrees or disagrees with these statements: Ⅲ Teaching health literacy to help consumers negotiate the health care system is a superficial fix (or ''too little too late'') because it fails to address the larger problems underlying the US health care system (such as 45 million uninsured Americans) Ⅲ Paradoxically, health literacy blames the victim for a deeper sociocultural problem Ⅲ Since health literacy efforts have not been generated at a grassroots level, they appear to be culturally paternalistic Hence, factor 1 can be characterized as supportive of health literacy's potential to improve patient cognitions and the clinical interactions between providers and patients. But factor 1 is less enthusiastic about other venues to initiate health literacy.

Factor 2: classroom education-oriented critics of health literacy initiatives.
In contrast with factors 1 and 3, factor 2 is interested in classroom education as a venue to advance health literacy initiatives. Unlike the other factors, factor 2 agrees that: Ⅲ The primary intent of health literacy initiatives should be to help primary or secondary school educators teach health literacy as a basic subject (e.g., such as English and math) Ⅲ Poor K-12 health education programs are the major barrier to improving health literacy In terms of what should be taught, factor 2 agrees or slightly agrees with the following statements: Ⅲ Health literacy initiatives should help more people of all ages understand how the human body functions rather than focusing on lay translations of medical terms and vocabulary Ⅲ Health literacy should conceptually encompass how patient misunderstandings and confusion add hidden costs to the nation's health care delivery system The latter seems to be both a rationale for classroom health literacy initiatives as well as possible subjects for instruction.
In contrast to factors 1 and 3, factor 2 also believes that public health intervention models provide a partially acceptable conceptual framework to advance health literacy activities and initiatives. For example, factor 2 strongly agrees or agrees that: Ⅲ Health literacy should conceptually shift its focus from improving comprehension skills to urging disease prevention and early detection Ⅲ The primary intent of health literacy initiatives should better document the causal pathway of how poor literacy affects health Among other perceptual differences with factor 1, factor 2 has reservations about whether patients and patient cognitive skills will be a beneficiary from health literacy initiatives. For example, factor 2 strongly disagrees, disagrees, or slightly disagrees with these statements: Ⅲ The primary intent of health literacy initiatives is to empower patients Ⅲ The primary intent of health literacy initiatives should be to improve the ability of consumers/patients/caregivers to think critically about the health information they receive Ⅲ Health literacy initiatives should focus on helping patients, caregivers, and consumers better navigate the health care delivery system Ⅲ The concept of health literacy should be expanded to include an emphasis on consumer skills to access health services and to engage in patient advocacy Ⅲ The concept of health literacy needs to better encompass the stress (even the panic) people feel when they need to know and then are left to seek health information Ⅲ The primary intent of health literacy initiatives should be to improve the ability of consumers/patients/caregivers to think critically about the health information they receive Ⅲ The primary intent of health literacy initiatives is to help patients better understand what a physician tells them as well as prescription information Instead, factor 2 seems to believe the impact of health literacy on patient-provider communication (and clinical interaction) may benefit physicians. Unlike factors 1 and 3, factor 2 strongly and slightly agrees that: Ⅲ The primary intent of health literacy initiatives is to improve patient adherence with physicians' instructions Ⅲ Health literacy efforts are a preemptive strike designed to reduce the liabilities of physicians, hospitals, and insurance carriers when patients do not understand the information presented to them In further contrast with factors 1 and 3, factor 2 slightly disagrees that: Ⅲ Improving health literacy will first result in patients who are more discerning about the medical advice they receive The latter three responses suggest that, to factor 2, health literacy initiatives may not have a benign impact on patients and consumer clinical interactions. To factor 2, physicians, hospitals, and insurance companies may benefit from health literacy initiatives. As a result, unlike factor 1, factor 2's responses suggest they are less comfortable with clinical settings as a primary venue to launch health literacy initiatives.
In contrast with factor 3, factor 2 also is not enthusiastic about health literacy initiatives in personal educational settings. Unlike factor 3, factor 2 strongly disagrees or disagrees that: Ⅲ The concept of health literacy should focus on a consumer's skill to interpret media messages, enable people to look for (and assess whom to ask for) more health information Ⅲ The primary intent of health literacy initiatives should be to improve the ability of consumers/patients/caregivers to think critically about the health information they receive Ⅲ The concept of health literacy should be expanded to include an emphasis on consumer skills to access health services and to engage in patient advocacy Finally, in contrast to both other factors, factor 2 is more skeptical about the long-range social consequences of health literacy initiatives. Factor 2 uniquely disagrees that ''Elevating the nation's health literacy is a vital step to improving the quality of health care and health outcomes.'' Factor 2 also uniquely either agrees or slightly agrees with these statements: Ⅲ Writing health materials at a sixth-grade level dilutes the quality of needed health information for all consumers Ⅲ Assigning people into marginal and low literacy categories sinks the cultural acceptability of future health literacy campaigns among the very audiences for whom they are intended Hence, factor 2 reflects a more critical perspective about health literacy than the other factors. But factor 2 believes health education is important and can be advanced in classroom educational settings.
Factor 3: personal education perspective. Factor 3 agrees with some broad statements that underscore the importance for consumers and patients to be better informed about health and medicine. For example, factor 3 (and factor 1) agree with these statements: Ⅲ Health literacy initiatives should focus on helping patients, caregivers, and consumers better navigate the health care delivery system Ⅲ The primary intent of health literacy initiatives is to empower patients Ⅲ The concept of health literacy needs to better encompass the stress (even the panic) people feel when they need to know and then are left to seek health information These responses suggest that factor 3 is mindful of the emotional dynamics that occur when patients and caregivers receive a clinical diagnosis and need health information. But, unlike factor 1 and factor 2, factor 3 uniquely strongly agrees, agrees, or slightly agrees with these statements: Ⅲ The concept of health literacy should focus on a consumer's skill to interpret media messages, enable people to look for (and assess whom to ask for) more health information Ⅲ The concept of health literacy should be expanded to include an emphasis on consumer skills to access health services and to engage in patient advocacy Ⅲ Foremost, health literacy should be conceptually conceived as improving a consumer's basic medical knowledge Factor 3 also is the only group to strongly agree that ''The primary intent of health literacy initiatives should be to improve the ability of consumers/patients/caregivers to think critically about the health information they receive.'' In short, while broad consumer, patient, and caregiver health education are important to factor 3, they are the group most interested in encouraging critical thinking, encouraging patient advocacy, and promoting basic medical knowledge. Factor 3 also uniquely agrees that helping consumers use the mass media to obtain health information should be an integral part of an approach to improve the public's health literacy.
These responses suggest that factor 3 is: (a) attentive to individual learning and (b) distinctly interested in what occurs when people use the mass media or inquire about health in settings without professional assistance. Moreover, by their relative lack of enthusiasm for clinical and classroom settings, factor 3's constituents seem to prefer to advance individual learning in situations where informal education resources are available. The latter is evidenced by factor 3's lack of enthusiasm for health literacy initiatives centered on clinical or classroom education settings.
For example, in contrast to factor 1, factor 3 is much less enthusiastic that ''The primary intent of health literacy initiatives is to help patients better understand what a physician tells them as well as prescription information.'' Similarly, unlike factor 1's stronger endorsement, factor 3 only slightly agrees with these statements: Ⅲ Improving health literacy will first result in patients who are more discerning about the medical advice they receive Ⅲ The primary goal for health literacy initiatives should be to increase the use of plain language in all consumer instructions and communication about health As a result, factor 3 does not seem to be an unequivocal supporter of focusing health literacy initiatives on clinical settings.
Unlike factor 2, factor 3 slightly disagrees with these statements: Ⅲ Poor K-12 health education programs are the major barrier to improving health literacy Ⅲ The primary intent of health literacy initiatives should be to help primary or secondary school educators teach health literacy as a basic subject (e.g., such as English and math) Hence, factor 3 has reservations about formal classroom educational settings, which contrast with factor 2's endorsement.
Overall, factor 3 is characterized by a distinctive attitude about the best setting to encourage personal knowledge and improve health literacy. While factor 3 agrees that improving the nation's health literacy is an important endeavor, the preferred setting to help people learn about medicine seems to be one that facilitates a more private, less formal, health information-seeking mode-such as when one seeks reading materials, browses the Internet for health Websites, or picks up information about medicine from the mass media versus a classroom or a physician's office setting. Factor 3's interests seem to be to help consumers learn when they are not necessarily surrounded by health education professionals or providers. Yet, factor 3 supports enabling consumers to discriminate among available health sources and materials.

DISCUSSION AND CONCLUSIONS
The findings suggest the respondents observe health literacy issues through different perceptual prisms. Factor 1 is optimistic and supportive of health literacy's transformative sociocultural and professional potential, if clinical settings become a launching point for health literacy activities. Factor 2 is less optimistic about health literacy's potential to improve clinical or patient outcomes and prefers to focus health literacy initiatives on classroom education settings. Factor 3 supports improving the nation's health literacy but is more inclined to support health literacy initiatives when people privately interact with materials.
The three factors disagree about the appropriate venue to launch successful health literacy efforts. While factor 1 believes primary care services represent optimal venues to focus health literacy efforts, factor 2 partially rejects this approach and places more confidence in classroom educational settings. Factor 3 disagrees with both factor 1 and factor 2 and seems to believe that health literacy initiatives may be more successful if they are focused on less formal, personal educational settings.
Although the findings and their implications are limited to the study's respondents, the study suggests that an educational setting makes a difference in how health care professionals perceive health literacy issues. The differences in health literacy attitudes also may reflect different educational priorities and preferences among health professionals, even if they broadly support the idea that improving the public's health literacy is important.
For each factor, the focal point regarding the appropriate venue to initiate health literacy initiatives also is somewhat mutually exclusive. While each factor supports its perspective about an appropriate educational setting, the findings suggest the other two options are less acceptable. For example, Table 1 reveals that factor 1 (which supports clinical settings to initiate health literacy activities) simultaneously fails to support classroom educational and personal educational settings as the most appropriate venue to initiate health literacy activities. Similarly, the other two factors reinforce their preferred educational settings and either disagree or are less enthusiastic about the other two.
More broadly, the acceptance by each factor of one perspective and partial lack of acceptance of other options suggests that the educational venues to initiate health literacy activities might be a source of future dissonance for some of the health care professionals who participated in this study.
The importance of these differences becomes evident if they are extrapolated to future health policy decisions, such as prioritizing financial and human resources to support health literacy initiatives. While the respondents were not asked how they would respond to prioritizing financial and human resources to support health literacy initiatives in clinical, formal classroom, or personal educational settings, the findings suggest each of the three factors would support deci-sions that reinforce their perspective. However, the findings imply that all three factors might hesitate about proposals to invest equally in clinical, education, or personal educational settings or invest unequally, if one factor's vested interest is funded inequitably at the expense of the other two.
Accordingly, the aggregate factor scores suggest the broader acceptance of health literacy initiatives might run into some professional resistance among the surveyed people. Among the study's respondents, a dialogue about clinical, classroom, and personal educational settings to initiate health literacy efforts should be appropriate, informative, and lively.
The study also implies it may be important to assess how health professionals prioritize and characterize the role of a person who seeks health care information and health services. For example, the study suggests that if people who seek health care information or services are identified as ''patients,'' ''students,'' or ''consumers,'' the assignment may make a difference in how health literacy initiatives are perceived and prioritized. The study suggests that the categorization of people who seek health care information and services could be an underlying issue that impacts health policy and health literacy perceptions. Certainly, the study suggests it is important for future research to address role perceptions and the associations between role assignments with broader perspectives about public health and health literacy initiatives.
Independent of these categorizations, the study also implies that respondents think about health literacy in terms of its impact on individuals as well as its impact on educational settings. Taking the holistic perspective recently advanced by Shohet and Renaud [30], individuals and educational settings are emphasized when health care professionals demur or disagree with sociological issues such as: (a) the influence of sociocultural undercurrents on the social impact of health care initiatives and (b) the impact of negative attitudes about health care organizations as social institutions on the social acceptance of health policy initiatives. Indeed, Table 1 reveals that all three factors agreed with few of the statements that addressed the sociocultural undercurrents that might adversely influence the social acceptance of health literacy. Yet Shohet and Renaud [30] emphasize that the latter issues are important to the public's acceptance of health literacy policy initiatives, which implies they should be of greater concern to more respondents.
It should be noted that the thematic patterns that distinguished the three factors were not necessarily predictable from the issues raised by the survey's individual items or from the three-themed structure that underlay the instrument. This illustrates one of the primary reasons to use Q methodology: its capacity to explore and discover perceptions that are not necessarily anticipated by the investigator [20,23,25]. The study additionally illustrates that mailing lists provide a promising venue to explore interprofessional discourse.
The study has important limitations, including a small, nonrandom sample size and a skewed sample of medical librarians that restrict the generalizability of the findings. The use of Q methodology (which has a smaller number of respondents than most public opinion research) inhibits a researcher's ability to explore whether differences or associations among demographic characteristics regarding health literacy issues are statistically significant. The interpretation of the factor structures also was thematically based, and alternative explanations of the arrays reported in Table  1 may be possible.
In future research, it would be interesting to determine if medical librarians and public health practitioners have attitudinal differences and whether educational level and background, professional allegiance, gender, geography, ethnicity, and other standard demographic factors differentiate opinion within one or both groups. It would be interesting to explore if the study's factor structures (or psychographic segmentation among clinical, formal educational, and personal educational settings) are sustained if a larger number of respondents are surveyed.
It should be noted that thematic patterns in factor arrays often remain stable in follow-up research that uses the same Q instrument [20,21,25]. Indeed, the thematic patterns in factor 1 and factor 2 in the current study (n ϭ 1,836) were similar to two of the factors in the pilot study (18 respondents and n ϭ 648).
Further, the findings suggest that a future study that assesses how health professionals rank health literacy versus other health policy priorities might be instructive. As listed in Figure 1, some of the health policy priorities that are sometimes mentioned as embedded or aligned with health literacy include: addressing the health care needs of underserved audiences, reducing health disparities or improving the access and delivery of care to other underserved demographic groups, and many other topics. It remains uncertain and worth investigating if health literacy is seen as a comparative priority in an array of policy efforts that seek to remedy other pressing public health challenges.
Finally, the implications of the differences among the respondents yield a basis for an evolving, fluid, vigorous dialogue about health literacy issues among health professionals. One dialogue might address the relative benefits of launching health literacy initiatives primarily in clinical settings or whether efforts should be directed to classroom or personal educational venues. Another dialogue might address how the perceptions of health literacy among health care professionals and medical librarians are impacted by their characterization of the people they serve.
The study suggests that intraprofessional dialogues about the perception of health literacy initiatives might foster greater understandings among health professionals, which could be a catalyst to increasing support for health literacy initiatives.