Abstract
Research has consistently revealed gender differences in attitudes toward science and technology. One explanation is that women are more personally affected by particular technologies (e.g., biomedical interventions), so they consider them differently. However, not all women universally experience biomedical technologies. We use the concept of technological salience to address how differences in subjective implications of a technology might explain differences in women’s attitudes toward biotechnology. In a sample of U.S. women from the National Survey of Fertility Barriers, we examine how women with and without a biomedical barrier to fertility evaluate biotechnology for infertility, which, we argue, reflects differences in technological salience. For women with a biomedical barrier, various experiences, beliefs, and values impacted their attitudes; yet, most of these did not affect attitudes if women had not experienced a fertility barrier. Results suggest that technological salience contextualizes women’s attitudes toward these biotechnologies and may also have broader implications for other biotechnologies.
Introduction
Research has consistently revealed gender differences in public attitudes toward science and technology. Women are typically more skeptical than men across a wide range of issues, including reproductive interventions (Napolitano and Ogunseitan 1999). One explanation is that women are more likely to be personally affected by particular technologies (Bryant and Pini 2006; Nelkin 1981). Yet most studies fail to link attitudes to experiences (except see Simon 2011). Feminist theory has provided some basis for understanding the gender–technology relationship, although feminists have disagreed as to whether technology liberates or oppresses women (Lublin 1998; Wacjman 1991). Recent theorizing, however, is more ambivalent—emphasizing that differences among women contextualize their relationships to technology: Experiences are not universal (Thompson 2002). Yet such insights have not been integrated into studies of attitudes toward technology: Most still treat gender categories as homogenous (e.g., Hayes and Tariq 2000).
In this article, we contribute to a theoretical framework of gender and attitudes toward technology that is based on experiences, focusing on biotechnology for infertility. Drawing on insights from feminist and other scholarship on technology, as well as social-psychological literature on attitudes, we address how differences in women’s fertility status affect how they may differentially use personal experiences, beliefs, and values to evaluate this technology. For fertility status, we distinguish between women who have and have not experienced a biomedical barrier to fertility. This is a broader definition of fertility problems that includes medically defined infertility (12 months of unprotected [hetero]sex without conception), miscarriage, and other problems conceiving or carrying a pregnancy to term (e.g., chronic illness). We argue that having or not having a biomedical barrier to fertility is a major category of difference shaping women’s attitudes about these technologies: We expect that women who have experienced a barrier will use different criteria when assessing the technology because it has greater subjective implications for them. We develop these ideas more generally with the concept of technological salience and test this on a sample of U.S. women from the National Survey of Fertility Barriers.
Background: Gender and Technology
Feminist scholarship has increasingly theorized the gender–technology relationship, and studies of public attitudes toward science have addressed gender differences in appraisals of science and technology. Yet these do not often come together in the same research. In theorizing technology, feminist scholarship has historically been deeply divided over viewing it as inherently oppressive or liberating for women (Lublin 1998). For instance, some view reproductive technologies as a form of patriarchal medical control over women’s bodies (e.g., O’Riordan and Haran 2009), while others have emphasized how technology can give women more control over reproduction (Wacjman 1991). More recent turns in feminist theorizing, however, have become more ambivalent, focusing on how women’s relationships to technology may vary across different social groups and even for one woman across her life course (Lublin 1998; Thompson 2002). For example, Thompson (2005) observed how some women simultaneously experienced infertility treatment as objectifying and liberating—a phenomenon she described as “agency through objectification.”
Other, often nonfeminist, research has focused on differences in men’s and women’s attitudes toward science and technology. Women are typically more pessimistic than men across a range of issues, including the environment (McCright 2010), nuclear power (Freudenburg and Davidson 2007), and biotechnologies (Napolitano and Ogunseitan 1999; Simon 2010). Some point to gender socialization: Women are socialized to be less interested in and less informed about science and technology (Tenenbaum and Leaper 2003). This “deficit model” explanation posits that when people know more about science and technology, they will have more favorable attitudes (Allum et al. 2008). A recent study, however, found that women’s increasing knowledge of technology actually contributed to their greater pessimism; for men this relationship was in the opposite direction (Simon 2010). In other words, men and women may evaluate science and technology through divergent processes.
One explanation is that women approach science and technology through a unique process, reflecting gender-specific consequences. Because women’s bodies are more often involved in childbearing, they are more likely to be exposed to the consequences of science and technology (Bryant and Pini 2006; Mallory 2006). For example, “Many of the applications [of biotechnologies] towards human health issues will likely affect fetuses, mothers, and young children more than [other] members of society” (Napolitano and Ogunseitan 1999, 202). This suggests that certain technologies may have greater personal implications for women than for men.
Two key insights from feminist science and technology studies are relevant here. First, women are overrepresented as users rather than producers of technology, which reinforces gender inequality: Women are often defined as “passive beneficiaries,” whereas men are the bearers of technological knowledge (Bray 2007). Second, women’s bodies are viewed differently than men’s in relation to technology. As Martin (2001, 197) observed, science and medicine use “models [of women’s bodies] implying failed production, waste, decay, and breakdown.” Therefore, when technology becomes personally relevant, it may also define women’s bodies as problematic.
Yet these various insights and explanations still ignore how women’s experiences with and relationships to technology differ: Not all women relate to reproductive technologies in the same way simply because they are women. Although women’s bodies may be defined as problematic (Martin 2001), for some women this becomes inherently more personal. In the next section, we develop technological salience as an explanatory concept describing how women’s evaluations of technology may differ vis-à-vis their personal history and experiences.
Conceptual Framework
Technological Salience and Attitude Formation
We draw on social-psychological insights about attitude formation to begin filling in some conceptual gaps. Research on attitudes suggests that personal relevance of an attitude-object (i.e., the topic about which an attitude is being formed) may affect how an individual processes information and makes a social judgment (Liberman and Chaiken 1996; Sorrentino et al. 1988). When an attitude-object has high personal relevance, individuals process information about it in a more thoughtful, holistic way; when an object is of low relevance, individuals tend to rely on simple rules or heuristics to create judgments (Sorrentino et al. 1988). Personal relevance therefore may activate a different set of information to create one’s attitude. Regarding attitudes toward technology, this implies that when the technology is more personally relevant (e.g., biotechnology for women’s health), an individual will form judgments in different ways than when it is of low relevance (e.g., space technology).
Yet, it is not simply that a technology is perceived as more or less personally relevant; when a technology has further subjective implications for someone as a potential user, we expect they might be more thorough in their evaluations, drawing on various types of information in creating judgments. We describe this as technological salience—the subjective implications of a specific technology for one’s lived experience. When a technology becomes more salient, it likely sensitizes individuals to consider the implications of a technology from a wider range of perspectives (e.g., scientific, religious, or ethical), thinking about how the technology can affect their lives.
Biotechnology for infertility arguably becomes salient when a woman has experienced a biomedical barrier to fertility. This makes the technology more personally relevant because she might actually use it. We would expect, then, that women who have ever had a biomedical barrier to fertility would use different criteria in forming their attitudes compared to other women. Strickler made a similar observation, addressing how clinicians versus infertility patients evaluated treatments in fundamentally different ways: Clinicians viewed infertility in terms of “diseases, which lend themselves to medical cures” (1992, 113), while patients considered psychological and emotional consequences of treatments. Here, differences in personal histories with and relationships to this technology gave rise to different criteria by which it was evaluated. Women for whom reproductive technologies had greater personal implications drew on a more extensive set of subjective evaluative criteria than the clinicians who administered but did not directly experience them.
It is important to note here that the salience of infertility technologies is itself deeply gendered, shaped by gendered assumptions within medicine as an institution: Historically there has been a greater emphasis on infertility as a women’s issue and much less focus on male infertility (Sandelowski 1993). Women become patients not only by being infertile but also by belonging to an infertile couple: Treatments for male infertility are overwhelmingly applied to women’s bodies (Greil 2002). Therefore, technological salience is not gender-neutral: Medical conditions and technological solutions are socially produced and gender norms greatly influence whose bodies are subject to treatment (Bray 2007).
Experience, Beliefs, and Values
What are the different evaluative criteria that technological salience might activate? Research on attitudes points to personal experience with the attitude-object (both direct and indirect) as an important factor in attitude formation (Fazio, Eiser, and Shook 2004). Additionally, attitudes do not exist in isolation, but are typically built on other belief and value orientations (Stern et al. 1995). For our study, we suggest that three main constructs tap into these: reproductive experience and values, technological experience, and contextual beliefs and values.
Reproductive experience and values
Women’s reproductive histories and values are likely to be highly relevant in shaping their attitudes about reproductive technologies. Women differently contemplate pregnancy; have/have not experienced pregnancy; and place different levels of importance on motherhood. Because there is a general consciousness that fertility technologies are becoming more available and more widely used (Broekmans et al. 2007; Jade Martin 2010), we might expect that any women who are currently pregnant, have been pregnant, or are contemplating pregnancy may be more aware of these technologies. However, reproductive experiences and values may be more impactful for women with a biomedical barrier to fertility. If women have been or are planning to become pregnant and also have a biomedical barrier, then infertility technologies become more relevant compared to women who can conceive without intervention. We consider reproductive experiences and values as distinct from technological experiences (see following) because women may not have experience with infertility and related technologies, but their personal situations make reproductive technology in general more relevant to their lives.
Technological experience
Having personal experience with a particular technology or knowing others who have can be important forms of knowledge shaping attitudes about medical technology (Gabe and Calnan 1989). Women who have had a biomedical barrier to fertility can draw on personal experience if they have sought any medical help. Women can also draw on indirect knowledge from social networks. In addition, women may have more confidence in the technology if they know a friend or family member who has successfully conceived from it; on the other hand, they may evaluate it more negatively if they know someone who has had unsuccessful treatment. When technological salience is high, we expect that both direct and indirect technological experiences will predict attitudes because women might be more likely to draw on that information. In contrast, when technological salience is low, women might be less likely to use (indirect) experiences because the experiences and the technology have a weaker connection. Rather, we would expect them to rely more on dominant cultural interpretations of the technology (i.e., “simple rules” for forming attitudes; Sorrentino et al. 1988).
Contextual beliefs and values
Reproductive processes are heavily influenced by broader belief systems and values that individuals hold. These might be important predictors of attitudes when technological salience is high because we expect that women would be more likely to consider how the technology might be assimilated into their lives. Assisted reproduction raises many ethical concerns, which may make women cautious, especially if scientists are perceived as “playing God” (Kalfoglou et al. 2005). These are often rooted in religious beliefs, which are frequently at odds with science and technology, including reproductive interventions (Roudsari, Allan, and Smith 2007). Because most religions oppose some form of reproductive intervention, religiosity may be more important in shaping women’s attitudes. Women’s personal locus of control may also factor in: Women who have a greater sense of personal control over their health and wellness may be less likely to place confidence in (external) medical authorities (Kornelson 2005; Nelson 1983). Finally, reproduction and family are highly gendered institutions that call on broader belief systems about men and women. Given the male-dominated character of science, technology, and medicine (Hayes 2001), we expect that women who espouse more traditional gender attitudes may also be more inclined to accept traditionally construed patriarchal medical authority.
Social Location
A final element to consider is how social structural location affects attitudes (Kiecolt 1988). While we do not necessarily suspect that the importance of social location varies by level of technological salience, it is still essential to consider this for attitudes toward fertility technologies. The concept of stratified reproduction relates women’s social location to experiences of reproduction (Colen 1986; Rapp 2001). Depending on social class, race, age, and marital status, among other factors, women are differentially empowered and encouraged to have/not have children, which likely affects their attitudes about reproductive technologies. Both imagined and actual U.S. fertility patients tend to be middle-class, white, heterosexual, married women. Infertility services are concentrated in a private medical market, so resources such as (private) insurance and income are key to access (Bell 2009; King and Meyer 1997). White women are much more likely to seek infertility help than Black and Hispanic women (Greil et al., 2011; Stephen and Chandra 2000); this is often interpreted via the history of distrust between Black and Hispanic groups and the U.S. medical system (Dovidio et al. 2008). Asian women are more similar to white women in their attitudes; however, recent research (Greil et al., 2011) indicates they may have greater ethical concerns about treatments. Age is both biologically and culturally significant because of the natural decline in women’s fecundity (Broekmans et al. 2007; Friese, Becker, and Nachtigall 2006). Infertility technologies may resonate more with older women. However, they may also be more doubtful about the technology’s efficacy compared to younger women, who might take fertility technologies for granted. Finally, infertility has historically been defined and treated as a medical problem within a married couple (Sandelowski 1993)—typically excluding single and lesbian women; however, more recent technology for egg freezing targets single, career-driven women to “protect” their future fertility (Jade Martin 2010). Therefore, while being married or in a heterosexual relationship may prime women to be more aware of these technologies, single women are also increasingly targeted for particular procedures.
To summarize, prior research has found gender differences in attitudes toward science and technology but has tended to assume within-gender experiences are similar (e.g., gender socialization produces women’s pessimism or viewing all women’s bodies as implicated in childbearing). Drawing on feminist insights that women do not universally experience and relate to technology and on social-psychological literature on attitudes, we argue that technological salience may be a key, overlooked concept in theorizing women’s attitudes about technology. When a technology becomes salient, we expect this will activate a wide range of individual experiences, beliefs, and values as criteria to evaluate the technology: These same factors might be less likely to affect attitudes when technological salience is low.
Reproductive interventions are an ideal subject with which to test the technological salience hypothesis because they are potentially applicable to all women, yet only some women require these interventions. Biotechnologies for infertility have more direct implications for women with a biomedical barrier to fertility than for other women. We expect that reproductive experiences and values, direct and indirect experiences with the technology, and other contextual beliefs and values will be associated with attitudes toward biotechnology for women who have had a biomedical barrier to fertility but not for other women.
Method
Data
Data came from wave one of the National Survey of Fertility Barriers (NSFB)—a random-digit-dialing telephone survey that addresses social and psychological aspects of childbearing and fertility problems among a nationally representative sample of U.S. women and a subset of partners (see Johnson and White 2009 for methodology report). Interviews were completed with 4,712 women between September 2004 and December 2006. There was a 53% percent response rate to the initial screener and 37.2% percent response rate to the final survey. Although this is relatively low, a comparison of key demographic and reproductive measures with two other national surveys with high response rates (National Survey of Family Growth and Current Population Survey) showed similar distributions of race/ethnicity, age, marital status, SES, and several fertility variables, including ever pregnant, ever infertile, and ever sought infertility treatment. To be interviewed, women had to be ages 25 to 45 and respond to introductory questions about wanting additional children, having ever been pregnant, and having ever experienced fertility problems. Women who had already had one child, did not want any more children, and had never experienced fertility problems were undersampled (20 percent randomly selected for full interview). Census tracts with 40% percent or more racial/ethnic minorities were oversampled to facilitate analyses by race/ethnicity.
For this study, we divided the sample into two groups for comparison: women who have ever had a biomedical barrier to fertility and women who have never had a barrier. This was constructed from self-reports in response to several questions. There were three main barrier pathways: (1) reporting a pregnancy (intended or unintended) that took more than one year to conceive, while not using birth control; (2) reporting one or more miscarriages; and (3) reporting another type of medical problem, such as sterilization prior to having desired children. Because of the diversity of women’s experiences, we controlled for type of barrier in models for those women (described below). Because objective or medically defined measures of infertility do not necessarily address subjective experiences, we also included a subjective measure of fertility problems (described below). Lesbian women can also be defined as having a fertility barrier and are likely to use assisted reproductive technologies when desiring biological children (Agigian 2004); however, because only 45 respondents self-identified as lesbian, controlling for this was not feasible.
After listwise deletion, the samples contained 2,887 (89.6% percent of the original sample) and 1,367 (87% percent of the original sample) women for those with and without a biomedical barrier to fertility, respectively. We compared characteristics for which there were complete data for both the original and analytic samples. There were similar distributions for age, race, parental status, most reproductive characteristics, and most technological experience characteristics. This suggested that the analytic sample did not dramatically differ from the original NSFB sample. In analyses, we weighted to adjust for nonresponse and sampling design.
Measures
Outcome
The main outcome was women’s confidence in biotechnology for infertility. This was constructed from two ordinal items: (1) “Medical science can be a big help to women having trouble getting pregnant” and (2) “Women who have trouble getting pregnant would benefit from consulting a doctor” (polychoric correlation = .43 for women with a biomedical barrier and .57 for women without). Items had four response categories (strongly agree to strongly disagree). These were initially part of a three-item scale with a planned missing design. Respondents were randomly asked two-thirds of the items to reduce respondent burden without compromising concept validity (Johnson and White 2009). Items were missing completely at random, so they do not bias the results (Allison 2002). We used singly imputed versions of items (from the NSFB public data) to construct our measure. We used the two-item scale because the third item was substantively different: It measured the extent to which respondents believed women could have children in their late thirties and beyond, with the help of “modern medicine.” Because the topic of advanced maternal age is associated with numerous ethical and medical concerns (e.g., increased risk of birth defects and complications), we excluded this in favor of addressing more general attitudes about medical intervention for infertility.
Social location control variables
While our analyses focus on differences in experiences, values, and beliefs, these are likely to be correlated with social location. We controlled for age in years. We indicated marital status comparing married (= 1) to all other statuses (= 0). We used a dichotomous measure of race (white = 1; nonwhite = 0). Because women could report multiple racial categories, anyone who reported at least one nonwhite category was coded as nonwhite. In preliminary analyses, we analyzed four categories (Black, Hispanic, Asian, and white) but found no significant differences, so we present the simplified measure here.
We used five variables for socioeconomic status: health insurance (private = 1; else = 0), state-mandated insurance (living in state with mandated infertility coverage = 1; else = 0), family income (12 categories, <$5,000 to $100,000+), employment (full time = 1; else = 0), and education (years). We compare private health insurance to all other options because in the United States, infertility treatment is concentrated in a private medical market and private insurance is key to access (King and Meyer 1997). We included state-mandated insurance because, though not a personal characteristic, it potentially mediates between personal insurance and access.
Reproductive experiences and values
We included five measures to address women’s reproductive experiences and values. We gauged parental status (= 1 if biological or social parent; = 0 if neither). Social parenting included adoption, stepchildren, and formal/informal fostering. An exclusive measure of biological parenthood showed no substantive difference in preliminary analyses. We indicated whether women were currently pregnant or trying to get pregnant (= 1) compared to women who were not trying or ambivalent (= 0). We indicated whether a woman had ever been pregnant (= 1) or not (= 0). We indicated whether women would like a(nother) baby in the future (= 1 if “probably” or “definitely” yes; = 0 if “probably” or “definitely” no). We also included the Importance of Parenthood Scale, which was composed of the following: (1) Having children is important to feeling complete as a woman; (2) I always thought I would be a parent; (3) Life . . . more fulfilling with children; and (4) It is important for me to have children. Items had four response categories (“strongly disagree” to “strongly agree”). A high score reflected a greater value of parenthood (α = .72 for women with a biomedical barrier, .78 for women without).
For women with a biomedical barrier, we included additional measures of their experiences in an extended model. We coded a respondent as self-identifying as having fertility problems (= 1) if they answered “yes” or “maybe” to the following: (1) Do you think you have/have had/might have trouble getting pregnant? (2) Do you think you have/have had a fertility problem? We indicated whether problems occurred with a first pregnancy (= 1) versus subsequent pregnancies (= 0). We controlled for type of fertility barrier: 12-month infertility and trying to conceive (infertile with intent), 12-month infertility but not trying to conceive (infertile no intent), miscarriage only, other medical issues (e.g., surgery, illness, sterilization of self/partner, and desire to have a child). We separated infertility into the intent/no intent groups because these women often have very different reactions (Greil and McQuillan 2004).
Technological experience
Several variables measured women’s technological experiences. We indicated whether they knew any friends or family who had used medical help to conceive: (1) Have family or friends pursued medical help to get pregnant? (2) Did any have a baby as a result of [medical] help? We recoded these into three dichotomous variables: (1) did not know anyone; (2) know someone, they were unsuccessful; (3) know someone, they had a baby.
In an extended model only for women with a biomedical barrier, we included a measure assessing the level of medical help seeking: no help seeking (= 0), considered only (= 1), talked to a doctor (= 2), had evaluation/diagnosis (= 3), and had treatment for the problem (e.g., artificial insemination) (= 4). If women had treatment, we further indicated whether it was successful, meaning it resulted in a current pregnancy or live birth (= 1); or unsuccessful, meaning no pregnancy, miscarriage, or stillbirth (= 0).
Contextual beliefs and values
The final set of variables included various beliefs and values relevant to family and reproduction. We indicated women’s ethical concern about infertility treatments via a three-item scale asking about artificial insemination, donor insemination, and in vitro fertilization. Three response categories ranged from “no ethical problem” to “serious ethical problems” (α = .72 for women with a biomedical barrier, .77 for women without). We included a six-item scale for medical locus of control. This assessed a woman’s perception of self-control over her health and well-being with statements such as “If I am sick, my own behavior determines how soon I get well again.” Each item had four response categories ranging from “strongly disagree” to “strongly agree” (α = .72 for women with a biomedical barrier, .68 for women without). We included a four-item scale measuring religiosity. This assessed frequency of religious behaviors, and how close they felt to God (α = .52 for women with a biomedical barrier, .57 for women without). Each item had different response categories; these were standardized before scaling. Finally, we included a measure of gender beliefs that assessed attitudes about men and women having separate spheres of responsibility for work and family life. The item had four response categories, ranging from “strongly agree” to “strongly disagree”: a high score indicated more traditional attitudes.
Analytic Strategy
Our main hypothesis is that because biotechnology for infertility has more personal implications for women who have had a biomedical barrier to fertility, they will use different criteria for appraising it than other women. To test this, we split our sample by biomedical barrier status because we suspect not just that the coefficients between these two groups will be unequal (i.e., interaction effects) but that entirely different factors may be at work in terms of what is shaping their attitudes about infertility technologies. We estimated both linear and ordinal logistic regression models because our outcome of interest was ordinal with seven categories. Results were substantively similar, so we present linear results for ease of interpretation.
Results
Table 1 displays descriptive statistics by biomedical barrier status. Bivariate associations show significant differences across each construct for these two groups of women. Notably, women who have never had a biomedical barrier have significantly more confidence in biotechnology for infertility on average than women who have ever had a biomedical barrier. This supports suggestions in prior research that women may be more skeptical of science and technology when their bodies are the focus of intervention (Bryant and Pini 2006; Nelkin 1981). These groups are similar across only three characteristics: percentage who know someone that sought medical help for pregnancy and did not have a baby; percentage who know someone that sought medical help and had a baby; and average medical locus of control score. The series of differences here further justify splitting our analytic sample. These differences also suggest that biomedical barrier status is potentially associated with different life course trajectories: Women with a biomedical barrier are more likely to be mothers, more likely to have experienced pregnancy or are trying to become pregnant, rate motherhood as more important, are more religious, and have more traditional attitudes. They also rate lower across all of the SES variables on average. It is possible that these characteristics place women in a position to test their fertility earlier on and discover problems.
Descriptive Statistics by BMB Status a
Weighted by NSFB survey design and population weights.
Difference of means tests performed for ordinal and continuous variables, χ2 tests for categorical.
p < .05, **p < .01, ***p < .001.
Table 2 shows linear regression results only for women with a biomedical barrier. Each successive model adds variables measuring the contributions of each construct in our framework. Model 1 controls for social location. Race, education, and income are all positively associated with women’s confidence in biotechnology in model 1, but in model 5 only the effects for race and education hold. White women express greater confidence in this biotechnology than non-white women, and more educated women express greater confidence than less educated women. These are both in the expected direction, based on prior research finding a link between higher education levels and greater acceptance of science and technology (e.g., Hayes 2001; Simon 2010), and research suggesting that non-white women may be less trusting of the medical system or have greater ethical concerns about treatments (Dovidio et al. 2008; Greil et al., 2011).
Linear Regression Results for Women’s Attitudes toward Biotechnology for Infertility, BMB Only (n = 2,887) a
Weighted by NSFB survey design and population weights.
p < .05. **p < .01. ***p < .001.
Model 2 adds several variables capturing women’s reproductive experiences and values. On average, women who had ever been pregnant were less confident in biotechnology for infertility. Many of these women may have achieved pregnancy without medical help, which may have lessened their need for and commitment to reproductive interventions. Women who placed greater value on motherhood as a desired social role tended to express greater confidence in this technology; this suggests that when being a parent is important, women may be more accepting of medical intervention to alleviate fertility problems. This latter result was the only finding from this set of variables that held in model 5.
Model 3 adds indirect technological experience: women’s knowledge of the technology through the experiences of friends and family members (Gabe and Calnan 1989). Compared to women who had no indirect experience, women who knew someone that sought medical help and successfully had a baby were more confident in the technology. However, contrary to our expectations, knowing someone who had unsuccessful treatment did not affect women’s confidence in the technology.
Model 4 adds in a series of beliefs and values that are important in contextualizing reproduction and family issues. As was expected, ethical concern was negatively associated with women’s confidence in the technology; however, there was no effect of religiosity on women’s attitudes. Counter to our expectations, women with higher personal locus of control about their health expressed greater confidence on average than women with lower locus of control scores. In addition, although we expected women with more traditional gender attitudes to be more likely to put confidence in external, traditionally patriarchal medical authority, we found the opposite result here: On average, these women expressed lower confidence in biotechnology for infertility. This pattern of results holds in model 5 as well.
Model 5 adds biomedical barrier-specific variables to the constructs of “reproductive experiences and values” and “technological experience” to account for additional experiences of these women. The type of fertility barrier and the level of help seeking were the only additional factors related to women’s attitudes. Women who had experienced other medical problems as opposed to infertility while intending to conceive had more confidence in the technology on average; however, women who were infertile without intending to conceive and women who had miscarried were not significantly different from those who were “infertile with intent.” Women in the “other problems” category have diverse issues, including sterilization regret, and desiring but not attempting pregnancy because of surgery or illness; these situations may put them in a position to more positively evaluate the possibilities of this biotechnology than women who are unable to conceive or who miscarry. Finally, women who moved further along the different medical help-seeking levels had greater confidence in the technology. To some extent, this contradicts the explanation that women are more critical of science and technology if their bodies are the focus of interventions (Bryant and Pini 2006; Nelkin 1981): If this explanation were to hold, we would expect to see women have decreased confidence as they become more involved in medical interventions. Without longitudinal data, however, we cannot tell if confidence enables women to continue with medical help or if continued interaction with fertility technologies increases women’s confidence. For instance, women who have negative experiences with medical help-seeking may drop out at lower levels, or women’s commitment to and confidence in reproductive interventions may increase the further they become involved. We suggest that this finding deserves significant exploration in future research.
Incremental F tests for each of the different models provide evidence for the successive contributions of each set of variables in explaining women’s attitudes. Notably, each set of variables in models 1 to 4 significantly contributed to the explanatory power of the model. This supports our conceptual framework in that each of these main constructs is both theoretically and empirically important in explaining variation in women’s attitudes about biotechnology for infertility—at least, for women with a biomedical barrier.
Table 3 provides linear regression results for women who have not had a biomedical barrier to fertility. To compare the patterns of effects, these contain identical variables as those in models 1 to 4 for women who have had a biomedical barrier. Model 1 includes the social location variables. On average, women who are older and have higher education levels tend to have greater confidence in biotechnology for infertility; however, only the effect for education holds in model 4. This is, again, in the expected direction based on prior research. One additional variable shows an effect only in model 4: White women are less likely to express confidence in the technology than nonwhite women. This is both counterintuitive, given prior research findings on race, infertility treatment, and medicine in general (Dovidio et al. 2008; Greil et al., 2011), and counter to the findings mentioned earlier for the women who have had a biomedical barrier. One explanation is that white, middle-class women are used to being in control of their fertility (e.g., through contraception)—and therefore often “experience the failure of infertility more acutely” (Greil 1991)—so they may be less supportive of this technology until they actually experience infertility.
Linear Regression Results for Women’s Attitudes toward Biotechnology for Infertility, Non-BMB Only (n = 1367) a
Weighted by NSFB survey design and population weights.
p < .05. **p < .01. ***p < .001.
Model 2 includes women’s reproductive experiences and values. Women who want a(nother) baby express less confidence in this technology; however, those who place a greater importance on motherhood reported greater confidence in the technology. It is possible that these reflect differences in ideal versus real circumstances for women: Ideally they desire parenthood and would possibly use technology to become a parent, but the reality of specifically wanting a(nother) child may make them more considerate of the risks and benefits of the technology.
Models 3 and 4 include the variables measuring (indirect) technological experiences and contextual beliefs and values. Unlike the previous models for women who have had a biomedical barrier, none of these variables were associated with women’s attitudes about biotechnology for infertility if they had not experienced a biomedical barrier. The incremental F tests for each of the different models provide evidence that variables introduced in models 1 and 2 significantly contributed to the explanatory power of the model, but those introduced in models 3 and 4 did not. In other words, social location and reproductive experiences and values helped to explain variation in women’s attitudes, but technological experience, and contextual beliefs and values, did not.
Comparing the results by biomedical barrier status reveals a number of differences between these groups. For women who had a barrier, measures from each of the major set of variables were associated with their attitudes toward biotechnology for infertility: race, education, importance of parenthood, and technological experience, as well as a range of personal beliefs and values. Many of these same findings were not relevant for women who had not had a biomedical barrier to fertility. Additionally, although each set of variables significantly contributed to the explanatory power of the model for women who had experienced a biomedical barrier (except for the variables added in model 5), only social location and reproductive experiences and values provided significant contributions for women who had not experienced any barriers. This pattern of results generally supports our technological salience framework: The greater subjective implications for women with a biomedical barrier activate different sets of evaluative criteria for judging the technology.
Discussion and Conclusion
A review of the existing research on gender differences in attitudes toward science and technology suggests that the more personal implications of technologies might explain why women evaluate them differently than men (Bryant and Pini 2006; Napolitano and Ogunseitan 1999; Nelkin 1981). However, feminist theorizing on gender and technology has argued that not all women universally experience technology simply because they are women (Lublin 1998; Thompson 2002). We know that technology (particularly reproductive technology) is not gender-neutral because gender norms deeply affect how technology is produced and consumed (Bray 2007) and because medico-technological solutions typically use a model that views women’s bodies as inherently prone to failure (Martin 2001). However, this should not universally impact all women. Rather, for some women we might expect biotechnologies to become inherently more personal. Drawing on these insights, as well as on social-psychological literature on attitudes, we developed the explanation of technological salience: When a technology has subjective implications, a different set of evaluative criteria will be used that draws more on experiences, values, and beliefs than when personal implications are low. When a technology is salient, it is something that could be done to you. Therefore, we would expect the evaluation to be based more on personalized criteria rather than on more abstract knowledge (Sorrentino et al. 1988).
To test this hypothesis, we examined women’s attitudes toward biotechnology for infertility. Some of the women in our sample have had a biomedical barrier to fertility so the biotechnology in question has more personal relevance. For other women, however, it does not have the same implications. Based on our framework, we expected that experiences and values related to reproduction, experiences with reproductive technologies, and beliefs and values that contextualize reproduction would be significantly correlated with attitudes for women who have had a biomedical barrier to fertility, but that these would not necessarily factor in for other women.
The results presented here do support our expectations regarding how technological salience may shape attitudes toward technology. Using regression analysis, we found that measures from each of our constructs—social location, reproductive experiences and values, technological experience, and contextual beliefs and values—were associated with women’s attitudes toward biotechnology for infertility if they had experienced a biomedical barrier to fertility, but only a few measures from the first two constructs were associated with attitudes for women who had not experienced a barrier. This different pattern of results supports the conceptual utility of technological salience: Relevant experiences, beliefs, and values are important criteria for determining attitudes when these interventions could have direct personal relevance for women but not for the women who would not be directly implicated in such interventions.
Notably, although there was no significant difference between these groups of women in the proportion who knew someone that had sought infertility treatment and successfully had a baby, this affected attitudes only for women with a biomedical barrier. When women are potential users of biotechnologies, it is intuitive that they would consider these indirect experiences as part of their evaluative criteria, but that these would not have the same impact for other women. That is, when a technology is particularly salient to a woman, she becomes more attuned to others’ experiences because they have the potential to be her own; when the technology is not particularly relevant, it may be easier to overlook how it has affected others.
Three other relevant beliefs and values variables significantly predicted attitudes for women with barriers and not for women without barriers: ethical concerns about infertility treatments, medical locus of control, and gender attitudes. While it is reasonable to expect those with ethical objections to have more negative attitudes toward medicine, it is noteworthy that this did not predict attitudes for women without a biomedical barrier. It appears that ethical concern comes into play only for women who would be (or have been) the subjects of such interventions, as they are called on to potentially assess how a technology might be assimilated into their personal belief and value systems. For women with a biomedical barrier, those who feel more in control of their own health and well-being actually have more confidence in medicine to help achieve pregnancy than women who feel less in control. It may be that women who feel more in control of their health outcomes view medical technology as a tool to be used when needed instead of viewing it as an external authority that must be conceded to: Medical interventions may become an extension of their own agency in getting pregnant. This supports Greil’s (1991) finding that infertile couples may turn to medical treatments to regain a sense of control that is lost when they experience infertility. Finally, having more traditional gender attitudes about separate spheres for men and women was negatively associated with confidence in biotechnology for infertility for women with a biomedical barrier, but had no effect for the other women. We initially expected this relationship would be positive: Women with traditional gender attitudes might be more likely to accept external medical authority, and women with more progressive gender attitudes might be more critical of medical interventions that underscore the inadequacies of their bodies. We would expect this to be true especially of medical interventions to pregnancy and birth, in which women have traditionally been excluded from decision-making processes by their husbands and male physicians (Leavitt 2010). Yet, for women with a biomedical barrier, more progressive gender attitudes were associated with more confidence in the technology. An alternative interpretation is to view it in terms of the link between the increasing number of reproductive interventions and women’s reproductive autonomy—even if it is only a discursive link for some reproductive technologies, for example, egg freezing (Jade Martin 2010)—rather than as forfeiting control of one’s body to traditionally construed patriarchal medical authority.
Overall, these findings supported our technological salience framework because we would expect ethical concerns, locus of control, and gender attitudes to be important considerations for evaluating fertility technologies for women with a biomedical barrier to fertility. When a woman is in a position to be subjected to infertility technologies (i.e., high salience), she becomes more likely to consider how to incorporate them into her own relevant beliefs and values; when salience is low, these same beliefs and values are not necessarily activated because these women would be much less likely to consider how to incorporate a technology into their lives.
In an expanded model in Table 2, we added several variables that applied only to women with a biomedical barrier to fertility: self-identification as having a fertility problem, having problems with a first pregnancy attempt (vs. subsequent pregnancies), type of fertility barrier, extent of medical help seeking, and whether any treatment sought was successful or unsuccessful. Type of fertility barrier and extent of help seeking further predicted women’s attitudes. Women who had experienced other medical problems as opposed to infertility while trying to conceive had more confidence in the technology on average. These women have more diverse medical issues that do not necessarily affect their ability to conceive—even though they affect their ability to have children. We suggested that this might cause them to more positively evaluate the possibilities of medical technologies than women who are specifically unable to conceive (medically defined infertility) or who miscarry. Women who moved further along the different medical help-seeking levels had more positive attitudes toward medical technology for helping infertile women achieve pregnancy, but we cannot disentangle this relationship without longitudinal data: We cannot tell if confidence in technology enables these women to pursue medical help or if continued interaction with medical help seeking increases women’s confidence in this technology. How women’s broader health beliefs and use of medical care play a role in their biotechnology attitudes deserves exploration in future research.
This study does have some limitations. First, we used cross-sectional data. Longitudinal data would provide a more comprehensive analysis of how experiences shape attitudes about technology; in particular, it is crucial to be able to disentangle the timing of attitude formation in relation to technological experience. Second, we focused on only one type of technology here, so it is an open question whether technological salience can be transferred to other technologies. We suspect that it may work for other reproductive technologies and food technologies, but perhaps not for technologies that are rarely accessed directly by the public (e.g., space exploration) or that do not have bodily implications. Another limitation that urges caution in interpretation is the relatively small amount of variance explained. The full model accounts for only about 10 percent of the variation in the dependent variable for those with a biomedical barrier, and about 8 percent for those without barriers. Future research should build on our conceptual framework to further flesh out this model. Another limitation is the relatively small sample size (n = 1,367 for women with a biomedical barrier). Because significance tests are a function of sample size, it is possible that we have rejected the possibility that there are correlations between the experiences, beliefs, and values variables and confidence in infertility technologies for this group of women when, in fact, there may be. Ideally, further research would perform similar analyses, but with a larger sample. A related sample issue is that NSFB respondents may not represent all U.S. women, even though many characteristics matched other surveys with much higher response rates. For example, women who have had particularly distressing experiences with infertility may have refused to participate because of the psychological toll it might take to relive past experiences. Thus, the findings here should be interpreted with caution. Finally, we have focused here only on women in order to isolate specific, relevant experiences and to critique prior research for its homogenous interpretation of gender categories. Future research should also address differences in how couples experience various reproductive processes, in order to assess how asymmetric personal involvement may differentially shape partners’ attitudes.
Despite limitations, this study has made several contributions to theories of gender and technology. First, it supports arguments made by scholars who have suggested that (gender) differences in attitude about science and technology may stem from different personal implications of the technologies in question (e.g., Hayes 2001; Simon 2010, 2011). Our work strengthens this explanation because we find technological salience to be an important contextualizing factor among women: We add to this literature further by acknowledging that gendered processes, including relationships to technology, cannot be conceptualized as universal among women (or men). Our work also has implications for feminist theories of reproduction (e.g., Strickler 1992; Wikler 1986) because it provides insight into how women approach reproductive technologies differently when “reproduction becomes problematic” (Rapp 2001, 466). Feminist theory has tended to focus on the implications of reproductive technologies for the women who might use them, but has paid less attention to how fertile and infertile women evaluate these technologies differently. Our research suggests that feminist theorizing should incorporate differences in technological salience among women into analyses of reproductive technologies. This line of inquiry should be particularly fruitful for exploring intragender differences in technology attitudes and experiences.
Footnotes
We would like to thank Pat Rafail and the anonymous reviewers for their comments on earlier versions of this manuscript.