Measuring the level of social support using latent class analysis

Different instruments have been used to measure social support in epidemiological studies of which the most widely used is the Medical Outcomes Study Social Support Scale (SSS-MOS). However, these studies lack measures of the level of social support on health risks. We used latent class analysis (LCA) to distinguish subgroups with different levels of perceived social support and tested the consistency of these subgroups by their associations with the prevalence of Common Mental Disorders (CMD). This is a cross-sectional study of 1013 mothers living in the city of Salvador, Brazil in which psychosocial data were collected through home visits using the SSS-MOS and the Self Reporting Questionnaire-20. For each dimension of social support analysed here, we selected models with two classes using LCA. Multivariate logistic regression models were used to estimate the association between participants’ perceived social support and the prevalence of CMD to verify the consistency of the groups defined by LCA. There was a clear difference in the reporting of perceived social support between those classified as high or low using LCA. The probability of perceiving several types of social support was lower in the subgroup classified as low level of social support (13.7–59.8%), and it was much higher in the group classified as high level of social support (84.3–98%). A greater prevalence of CMD was found among mothers with lower levels of social support. LCA seems to be a useful tool to improve measurement of perceived social support by separation into two levels in which the lower level is associated with an increased prevalence of CMD.     

Improving ability measurement in surveys by following the principles of IRT: The Wordsum vocabulary test in the General Social Survey

Survey researchers often administer batteries of questions to measure respondents’ abilities, but these batteries are not always designed in keeping with the principles of optimal test construction. This paper illustrates one instance in which following these principles can improve a measurement tool used widely in the social and behavioral sciences: the GSS’s vocabulary test called “Wordsum”. This ten-item test is composed of very difficult items and very easy items, and item response theory (IRT) suggests that the omission of moderately difficult items is likely to have handicapped Wordsum’s effectiveness. Analyses of data from national samples of thousands of American adults show that after adding four moderately difficult items to create a 14-item battery, “Wordsumplus” (1) outperformed the original battery in terms of quality indicators suggested by classical test theory; (2) reduced the standard error of IRT ability estimates in the middle of the latent ability dimension; and (3) exhibited higher concurrent validity. These findings show how to improve Wordsum and suggest that analysts should use a score based on all 14 items instead of using the summary score provided by the GSS, which is based on only the original 10 items. These results also show more generally how surveys measuring abilities (and other constructs) can benefit from careful application of insights from the contemporary educational testing literature.     

New options for national population surveys: The implications of internet and smartphone coverage

Challenges to survey data collection have increased the costs of social research via face-to-face surveys so much that it may become extremely difficult for social scientists to continue using these methods. A key drawback to less expensive Internet-based alternatives is the threat of biased results from coverage errors in survey data. The rise of Internet-enabled smartphones presents an opportunity to re-examine the issue of Internet coverage for surveys and its implications for coverage bias. Two questions (on Internet access and smartphone ownership) were added to the National Survey of Family Growth (NSFG), a U.S. national probability survey of women and men age 15–44, using a continuous sample design. We examine 16 quarters (4 years) of data, from September 2012 to August 2016.Overall, we estimate that 82.9% of the target NSFG population has Internet access, and 81.6% has a smartphone. Combined, this means that about 90.7% of U.S. residents age 15–44 have Internet access, via either traditional devices or a smartphone. We find some evidence of compensatory coverage when looking at key race/ethnicity and age subgroups. For instance, while Black teens (15–18) have the lowest estimated rate of Internet access (81.9%) and the lowest rate of smartphone usage (72.6%), an estimated 88.0% of this subgroup has some form of Internet access.We also examine the socio-demographic correlates of Internet and smartphone coverage, separately and combined, as indicators of technology access in this population. In addition, we look at the effect of differential coverage on key estimates produced by the NSFG, related to fertility, family formation, and sexual activity. While this does not address nonresponse or measurement biases that may differ for alternative modes, our paper has implications for possible coverage biases that may arise when switching to a Web-based mode of data collection, either for follow-up surveys or to replace the main face-to-face data collection.     

Estimating chance reproducibility in Guttman Scaling

A definition of reproducibility in Guttman Scaling and two chance measures of reproducibility are suggested. The first measure assumes that the items are independent. The second method assumes nonindependent items and fits respondent and item margins by an iterative method used in fitting log-linear models. Chance reproducibility is conceptualized in terms of assumptions about respondent variability.     

Growing pains or appreciable gains? Latent classes of neighborhood change,and consequences for crime in Southern California neighborhoods

This study explored the dynamic nature of neighborhoods using a relatively novel approach and data source. By using a nonparametric holistic approach of neighborhood change based on latent class analysis (LCA), we have explored how changes in the socio-demographic characteristics of residents, as well as home improvement and refinance activity by residents, are related to changes in neighborhood crime over a decade. Utilizing annual home mortgage loan data in the city of Los Angeles from the years 2000–2010, we 1) conducted principle components factor analyses using measures of residential in-migration and home investment activities; 2) estimated LCA models to identify classes of neighborhoods that shared common patterns of change over the decade; 3) described these 11 classes; 4) estimated change-score regression models to assess the relationship of these classes with changing crime rates. The analyses detected six broad types of neighborhood change: 1) stability; 2) urban investors; 3) higher-income home buyers; 4) in-mover oscillating; 5) oscillating refinance; 6) mixed-trait. The study describes the characteristics of each of these classes, and how they are related to changes in crime rates over the decade.     

Rasch measurement and differences between women and men in self-esteem

Most researchers who investigate sociological influences on emotional well-being or distress employ correlational methods as justification for the scales used as measures of distress (factor analysis) and to describe the effects of various factors on distress (regression). This paper suggests that log-linear and related models offer certain advantages over the linear model which have largely been ignored. An analysis of the cross-classification of items included in a widely used measure of self-esteem suggests that one of the items must be eliminated to approximately achieve measurement as defined by Rasch. The resultant latent variable shows no difference in self-esteem between women and men. An analysis of the additive index which includes all of the items, however, reveals that women have lower self-esteem than do men. This apparent difference results solely from a difference on the discrepant item: women are less likely than men to report that they often, rather than sometimes or rarely, can do things as well as most other people.     

General population estimates of the association between college experience and the odds of forced intercourse

Sexual assault on college campuses is a pervasive problem, recently receiving increased scientific and policy attention. However, the high focus on college student experience ignores general population prevalence, trends, and differences between those with college experience and those without. We analyze measures from the National Survey of Family Growth (NSFG) to provide a general population view of experiences with forced intercourse. Forced intercourse is a common experience in the U.S. population, has remained stable in recent years, and varies greatly by gender, age and race. The odds of forced intercourse are also significantly higher among those with less than four years of college. This ubiquitous public health problem is not limited to college campuses. Measures from the NSFG are an important resource for understanding population rates of (and trends in) forced intercourse, providing information to guide interventions and better target scientific investigation.     

Form effect in the measurement of feeling states

This study investigates the relation between the responses given to survey items and the manner in which those items are given. Results are based on responses of a random sample of 1522 American adults to 202 items expressing social life feelings (SLFs). SLF items were selected from a domain of over 1000 such items appearing in over 100 scales used in American sociology during the last 50 years. Three different methods of administration were used: self-administration, interview, and card-sort. Each method was applied to each item about 500 times, and to each respondent approximately 67 times. The analysis consisted of comparing mean scores and missing response rates for items across methods, investigating the characteristics of items subject to form effects, investigating the possibility that form effects were spurious and might disappear after controlling on background of respondent, and investigating the presence of form effects in summary scale scores based on several or more items. Results indicate that for most items, patterns of responding and techniques of testing were statistically independent.     

Looking for homogeneous groups of respondents and nonrespondents using latent class analysis

Concerns about decreasing response rates have led to many nonresponse studies. Unfortunately, usually only a limited amount of data is available on nonrespondents. Linking administrative data makes information available on all the sampled units including nonrespondents. Statistics Netherlands has constructed a database in which several administrative records and several surveys are connected. In this explorative study we use this dataset to look for homogenous groups of respondents and nonrespondents in survey research using latent class cluster analysis. Latent class techniques can provide insight into the problem of selective nonresponse and currently under-represented societal groups. We identify four different latent classes in a recent Dutch survey and replicate the findings for the same survey conducted 4 years later. Two of the types of sampled units have above-average response rates and the other two have below-average response rates. We also evaluate various latent class models with other response outcomes to gain insight into the contact and cooperation process.     

Detecting Discrimination: Analyzing Racial Disparities in Public Contracting

InCity of Richmond v. J. A. Croson Co.(1989), the Supreme Court established strict scrutiny as the standard applicable to affirmative action programs which set aside quotas of public contracts for minority-owned businesses, andAderand v. Pena(1995) extended the strict scrutiny standard to federal programs. Although the requirements of these decisions clearly require multivariate statistical analysis, most “disparity studies” have used a univariate comparison between the expected and the observed shares of contracts going to minority-owned firms. We examine four statistical methods—ordinary least square multiple regression, logit and tobit models, and a multivariate procedure for comparing expected and observed outcomes. Because no data are presently available at the level of specificity required byCroson,we constructed synthetic data sets to represent typical variations among large U.S. cities. Applying the statistical methods to each data set allows evaluation of the extent to which each method is able to both remove spurious and detect valid estimates of racial disparity when relevant control variables are added. Findings: (a) All four models removed apparent disparities which, although significant in univariate analysis, were known to be spurious. (b) Tobit and logit models, whose underlying assumptions better fit the nature of public contracting data, provided more accurate and more sensitive estimates than OLS regression. (c) Comparison of expected and observed outcomes within categories of control variables yielded results very similar to logit and tobit models and, because of the nature of the comparison specified inCroson,produced slightly more sensitive probability estimates.     

Best news yet on the six-factor model of well-being

This study provides a response to Springer and Hauser’s claim that the Ryff Scales of Psychological Well-Being do not comprise six dimensions. We show that their analyses support the theory-guided six-factor model, although their interpretations of the data reveal a lack of understanding of the construct-oriented approach to personality assessment. We also review evidence from five categories of studies (i.e., factorial validity, psychological correlates, sociodemographic correlates, biological correlates, and intervention studies) that document the distinctiveness of the six dimensions of well-being, including among subscales that are highly correlated. We conclude with an examination of the methodological corrections (negative items, adjacent items, and redundant items) employed by Springer and Hauser, finding all to be problematic, either in terms of scientific rationale or method of implementation, or both. Correlations among latent constructs resulting from these analyses are thus highly questionable.     

Symmetry,quasi-symmetry,and marginal homogeneity on the latent level

(Quasi-)symmetry models can be applied to panel and trend data in many ways, some of which are seemingly totally unrelated to the concept of (quasi-)symmetry. Log-linear models are very well suited to estimate the amount of (quasi-)symmetry in a square frequency table. Log-linear models with latent variables make it possible to assess the (quasi-)symmetrical nature of the multivariate distributions of the latent variables, thereby correcting for the unreliability in the measurement of the manifest variables. Latent (quasi-)symmetry models can be applied by introducing a slight extension of Goodman's well-known version of the EM-algorithm for the estimation and testing of latent structure models.     

Using geocoded survey data to improve the accuracy of multilevel small area synthetic estimates

This paper examines the secondary data requirements for multilevel small area synthetic estimation (ML-SASE). This research method uses secondary survey data sets as source data for statistical models. The parameters of these models are used to generate data for small areas. The paper assesses the impact of knowing the geographical location of survey respondents on the accuracy of estimates, moving beyond debating the generic merits of geocoded social survey datasets to examine quantitatively the hypothesis that knowing the approximate location of respondents can improve the accuracy of the resultant estimates. Four sets of synthetic estimates are generated to predict expected levels of limiting long term illnesses using different levels of knowledge about respondent location. The estimates were compared to comprehensive census data on limiting long term illness (LLTI). Estimates based on fully geocoded data were more accurate than estimates based on data that did not include geocodes.     

论《SPS协定》中的科学证据原则

科学证据原则是《SPS协定》的核心原则之一。分析科学证据原则的核心条款第2.2条和例外条款第5.7条是认识该原则的关键。在法律条款的适用上,应区分事实属性与法律适用两种情况,遵循先具体条款再原则条款的适用模式。争端解决机构对荷尔蒙案、鲑鱼案和品种测试案的裁决为理解这一原则提供了实践依据。     

Restrictive abortion laws, antiabortion attitudes and women’s contraceptive use

This study empirically examines the public and social policy question: Do state restrictive abortion laws affect the likelihood that women use more highly effective contraceptive methods? Using contraceptive use data from the Behavioral Risk Factor Surveillance System 2002 survey, the empirical results show that Medicaid Funding Restrictions, Informed Consent Laws, and Two-Visit Laws have no significant impact on adult women’s (ages 18-44, 18-24, 25-34, 35-44) use of highly effective contraceptive methods. A state’s antiabortion attitudes, which likely contribute to the enactment of restrictive abortion laws in a state, are a major factor in inducing greater use of highly effective contraceptive methods by adult women at-risk of an unintended pregnancy. The empirical findings remain robust for various population subgroups of adult women (i.e., married, single, employed, unemployed, with children, no children and college educated).     

The Republicanization of evangelical Protestants in the United States: An examination of the sources of political realignment

Although the association between evangelical Protestant and Republican affiliations is now a fundamental aspect of American politics, this was not the case as recently as the early 1980s. Following work on secular political realignment and the issue evolution model of partisan change, I use four decades of repeated cross-sectional survey data to examine the dynamic correlates of evangelical Protestant and Republican affiliations, and how these factors promote changes in partisanship. Results show that evangelical Protestants have become relatively more likely to attend religious services and to oppose homosexuality, abortion, and welfare spending. Period-specific mediation models show that opposition to abortion, homosexuality, and welfare spending have become more robust predictors of Republican affiliation. By the twenty-first century, differences in Republican affiliation between evangelical Protestants and other religious affiliates are fully mediated by views of homosexuality, abortion, and welfare spending; and differences in Republican affiliation between evangelicals and the religiously unaffiliated are substantially mediated by views of homosexuality, abortion, welfare spending, and military spending. These results further understanding of rapid changes in politico-religious alignments and the increasing importance of moral and cultural issues in American politics, which supports a culture wars depiction of the contemporary political landscape.     

The effects of religious contextual norms, structural constraints, and personal religiosity on abortion decisions

Researchers have established that individual religiosity influences abortion attitudes, and that abortion attitudes, in turn, shape abortion restrictions and access. Less clear is whether religion and abortion structural constraints influence abortion decisions. This study examines the several individual, contextual, and structural factors that could shape the abortion decisions of women who conceive before marriage. Special attention is given to the importance of academic aspirations and structural constraints, in contrast to religious beliefs and county religious context, for making an abortion decision. Hierarchical modeling techniques and two waves of data from the National Longitudinal Study of Adolescent Health (Add Health) are employed. Neither generic religiosity nor conservative Protestant religious context appear to influence women’s abortion decisions. Conversely, young women’s abortion decisions are shaped by academic ambition, identification with a conservative Protestant denomination, proximity to an abortion clinic and the level of public abortion funding in their county of residence.     

“公共支出均衡”的分析：中国样本

因为资源的稀缺性,公共财政支出的内部结构存在着均衡态,即公共支出均衡.不同于市场均衡,公共支出均衡意味着在政府公共支出规模约束下经济效率目标与公平产品目标的同时最大化.从均衡视角评估公共支出结构正是经济学对政府支出定位的应有之意.基于人类发展指数的较长时间序列相关数据的实证检验表明,我国公共支出中各支出项目对HDI的边际贡献率在时间序列上表现出显著的收敛趋势,这说明我国公共支出结构在长期中表现出均衡改进的特征.而各支出项目边际贡献率变迁趋势的差异性则表明依照均衡原则调整并优化我国公共支出结构是亟待之需.     

农业劳动力转移的动力机制及其对粮食安全的影响

通过构建C-D生产函数考察劳动及资本要素对粮食产量的影响,进而分析劳动的边际产量变化情况.在此基础上,结合劳动力的机会成本分析劳动力转移的动力机制及其对粮食安全的影响.结论：长期的高机会成本使粮食种植的劳动力投入持续减少,已对我国粮食生产产生了不利影响;而耕地的社会保障功能使其规模化生产难以实现,从而使通过资本替代劳动来提高粮食生产率变得困难,这造成了劳动力转移对我国粮食安全的潜在风险.建议：近期内采取措施提高种粮的比较收益以吸引部分青壮年留在农业;远期则需建立起有效的社会保障体系,从而消除土地的社会保障功能,以通过耕地的规模化生产来提高农业生产率,进而保证粮食安全.     

高校辅导员与大学生人际关系问卷的编制

目的：编制高校辅导员与大学生人际关系影响因素问卷。方法：通过对565名大学生调查数据进行探索性因素分析,检验问卷的信度和效度。结果：高校辅导员与大学生人际关系影响因素问卷包含20个题项,分为工作技能、社会评价、仪容仪表、理论素养、交往频率5个因子,转轴后5个共同因素可以解释的总变异量为57．962％,题项的负荷范围是0．621～0．851;总问卷的内部一致性系数为0．840,5个分问卷的内部一致性系数都在0．65以上。结论：问卷的因素结构清晰、信度和效度均较好。