Nonresponse in repeat population-based voluntary counseling and testing for HIV in rural Malawi

Despite the increasing number of population-based surveys in sub-Saharan Africa that provide testing and counseling for HIV over the past decade, understanding the nature of nonresponse in these surveys, especially panel HIV surveys, is still limited. This article uses longitudinal HIV data collected from rural Malawi in 2004 and 2006 to examine nonresponse in repeat population-based testing. It shows that nonresponse in repeat testing led to significant bias in the estimates of HIV prevalence and to inconsistent conclusions about the predictors of HIV status. In contrast, previous cross-sectional analyses found that nonresponse does not significantly bias the estimates of HIV prevalence. The difference in conclusions from cross-sectional and longitudinal analyses of nonresponse can be attributed to two factors: the different definitions of what constitutes nonresponse in both contexts, and the risk profiles of the missed populations. In particular, although refusal and temporary absence are the major sources ofnonresponse in the cross-sectional contexts, attrition attributable to mortality and out-migration are additional sources of nonresponse in repeat testing. Evidence shows that out-migrants have higher HIV prevalence than nonmigrants, which could account for significant bias in the estimates of prevalence among participants in both tests observed in this study.     

Victimization Among Female and Male Sexual Minority Status Groups: Evidence From the British Crime Survey 2007–2010

International surveys of victims show crime rates in England and Wales, including hate crimes, are among the highest in Europe. Nevertheless, sexual minority status is a less considered risk factor in general victimization research. This study used sexual minority status and sex to predict victimization across British Crime Surveys from 2007–2010. Logistic regression analyses showed sexual minority status groups were more likely than heterosexuals to be victimized from any and some specific crimes. However, bisexuals rather than lesbians or gay men were more consistently victimized, notably by sexual attacks and within the household. Implications for understanding victimization among these groups are discussed.     

A Multidimensional Dynamic Measure of Child Disadvantage: A Methodological Tool for Policymakers

This paper demonstrates the applicability of an innovative approach towards examining child disadvantage, using a holistic, dynamic measure that not only accounts for multiple sources of disadvantage but also for the recurrence and persistence of disadvantage throughout a child’s life. We analyse child disadvantage using two longitudinal surveys of the Australian child population, one of which is specific to Indigenous children, who experience notably higher rates of disadvantage. Among Australian children, we detect that poor body weight and bullying—representative of the broad dimensions of health and emotional wellbeing—should be of significant concern to policymakers. Among Indigenous children, housing conditions, schooling and exposure to risky behaviours stand out as areas of concern. By identifying the dimensions in which rates of child disadvantage are most severe, this methodological approach can help steer targeted policy actions.     

Refusal Bias in the Estimation of HIV Prevalence

In 2007, UNAIDS corrected estimates of global HIV prevalence downward from 40 million to 33 million based on a methodological shift from sentinel surveillance to population-based surveys. Since then, population-based surveys are considered the gold standard for estimating HIV prevalence. However, prevalence rates based on representative surveys may be biased because of nonresponse. This article investigates one potential source of nonresponse bias: refusal to participate in the HIV test. We use the identity of randomly assigned interviewers to identify the participation effect and estimate HIV prevalence rates corrected for unobservable characteristics with a Heckman selection model. The analysis is based on a survey of 1,992 individuals in urban Namibia, which included an HIV test. We find that the bias resulting from refusal is not significant for the overall sample. However, a detailed analysis using kernel density estimates shows that the bias is substantial for the younger and the poorer population. Nonparticipants in these subsamples are estimated to be three times more likely to be HIV-positive than participants. The difference is particularly pronounced for women. Prevalence rates that ignore this selection effect may be seriously biased for specific target groups, leading to misallocation of resources for prevention and treatment.     

Development of a social monitoring system at the local level: An experiment in the Southern Region of the Democratic Republic of Sudan

Subjective indicators of life quality are used frequently in the planning and decision making process. The limitations inherent in the use of subjective indicators often go unnoticed among those most likely to use such data in the context of planned change. We direct attention to a potential source of error in subjective evaluations of community services and facilities. Specifically, we utilize repeated measures to test for response bias that results from placing items in alternative sections of the same survey instrument. Evidence concerning the hypothesized presence of systematic bias was mixed. When overall service evaluation items were embedded within a section devoted exclusively to that service, considerable systematic bias was observed. The findings indicate that the choice of question format depends upon the ultimate use to which the data will be put.     

Sampling Weights for Analyses of Couple Data: Example of the Demographic and Health Surveys

In some surveys, women and men are interviewed separately in selected households, allowing matching of partner information and analyses of couples. Although individual sampling weights exist for men and women, sampling weights specific for couples are rarely derived. We present a method of estimating appropriate weights for couples that extends methods currently used in the Demographic and Health Surveys (DHS) for individual weights. To see how results vary, we analyze 1912 estimates (means; proportions; linear regression; and simple and multinomial logistic regression coefficients, and their standard errors) with couple data in each of 11 DHS surveys in which the couple weight could be derived. We used two measures of bias: absolute percentage difference from the value estimated with the couple weight and ratio of the absolute difference to the standard error using the couple weight. The latter shows greater bias for means and proportions, whereas the former and a combination of both measures show greater bias for regression coefficients. Comparing results using couple weights with published results using women’s weights for a logistic regression of couple contraceptive use in Turkey, we found that 6 of 27 coefficients had a bias above 5 %. On the other hand, a simulation of varying response rates (27 simulations) showed that median percentage bias in a logistic regression was less than 3 % for 17 of 18 coefficients. Two proxy couple weights that can be calculated in all DHS surveys perform considerably better than either male or female weights. We recommend that a couple weight be calculated and made available with couple data from such surveys.     

Rural-to-Urban Migration and Sexual Debut in Thailand

Migration from one’s parents’ home and sexual debut are common features of the transition to adulthood. Although many studies have described both of these features independently, few have examined the relationship between migration and sexual debut in a systematic manner. In this study, we explore this link for young adults in Thailand. With relatively high rates of internal migration, rapid modernization, a moderate HIV epidemic, and a declining average age of sexual debut, Thailand presents an instructive environment in which to examine migration and sexual debut. We use two waves of a longitudinal data set (2005 and 2007) that includes a subsample of young adults who migrated to urban areas during that period. We identify characteristics and behaviors associated with sexual debut and examine the role of migration on debut. Our approach reduces several common sources of bias that hamper existing work on both migration and sexual debut: (1) the longitudinal nature of the data enables us to examine the effects of characteristics that predate both behaviors of interest; (2) the survey on sexual behavior employed a technique that reduces response bias; and (3) we examine differences in debut by marital status. We find that migrants have a higher likelihood of sexual debut than nonmigrants.     

Sexual minority youth victimization and social support: the intersection of sexuality, gender, race, and victimization

In comparison to heterosexual youth, sexual minority youth are more likely to experience victimization. Multiple studies have connected anti-gay prejudice and anti-gay victimization to negative outcomes. Research shows that social support may protect sexual minorities from the harmful effects of anti-gay victimization. However, rates of victimization and the negative outcomes linked to sexual identity within the sexual minority community have been relatively unexplored. Using data from three years of statewide data from heterosexual and sexual minority adolescents in grades 9-12, this study examines victimization, substance use, suicidality, and access to social support by sexuality. Results indicate that sexual minority youth are at increased risk for victimization, substance use, suicidality, and social isolation compared to their heterosexual counterparts. Results also indicate that there is very little bivariate difference within the sexual minority community. Multivariate results indicate differences among sexual minorities' experiences with victimization and substance use.     

Mortality in Vietnam, 1979–1989

Little is known about past and present mortality in Vietnam, as the first official data on mortality have only recently become available from censuses taken in 1979 and 1989. Using these data, I estimate Vietnamese mortality during the intercensal period using two techniques that rely on age-specific growth rates from two successive age distributions. Intercensal emigration and differential completeness of census enumeration associated with massive outflows of refugees in the wake of the Vietnam War; population-redistribution policies, and a highly mobile population represent important sources of bias for the estimation of intercensal mortality. I incorporate several strategies to minimize bias from these sources and to select the method that is least sensitive to errors associated with them. Life expectancy at birth estimated for the 1979–1989 intercensal period is 61.4 years for males and 63.2 for females. These results suggest a trend of declining mortality between the 1970s and the 1980s and add solid empirical evidence to the debate over whether mortality in Vietnam has been deteriorating or improving.     

A probability model for census adjustment

The census can be adjusted using capture‐recapture techniques: capture in the census, recapture in a special Post Enumeration Survey (PES) done after the census. The population is estimated using the Dual System Estimator (DSE). Estimates are made separately for demographic groups called post strata; adjustment factors are then applied to these demographic groups within small geographic areas. We offer a probability model for this process, in which several sources of error can be distinguished. In this model, correlation bias arises from behavioral differences between persons counted in the census and persons missed by the census. The first group may on the whole be more likely to respond to the PES: if so, the DSE will be systematically too low, and that is an example of correlation bias. Correlation bias is distinguished from heterogeneity, which occurs if the census has a higher capture rate in some geographic areas than others. Finally, ratio estimator bias and variance are considered. The objective is to clarify the probabilistic foundations of the DSE, and the definitions of certain terms widely used in discussing that estimator.     

Change, continuity or cycles: The population turnaround in New South Wales

This study examines population flows from Sydney and other regions to perimetropolitan and coastal areas of New South Wales, the two main foci of the population turnaround since 1971. It uses census internal migration statistics for the five intercensal periods between 1971 and 1996, and estimated resident population statistics between 1997 and 2000. Fluctuating trends are described with respect to variations in age structures of migration flows and net migration gains by SLAs in coastal areas over time. Evidence of fluctuating trends is evaluated in relation to population structure change and local socio-economic multipliers in turnaround areas. The turnaround is far mor than a net migration gain from large metropolitan areas; it is also associated with interregional migration which avoids metropolitan areas, and which is at least in part environment- and amenity-related. The experience of some other countries, such as the USA where net migration reversals in population turn-around regions have occurred, has not been replicated in Australia. While elements of explanation for the complex cumulative causation process of the population turnaround in Australia are discussed, including the issues of fluctuating or cyclical trends, much more understanding of the economic and social factors involved is required.     

The Relationship History Calendar: Improving the Scope and Quality of Data on Youth Sexual Behavior

Most survey data on sexual activities are obtained via face-to-face interviews, which are prone to misreporting of socially unacceptable behaviors. Demographers have developed various private response methods to minimize social desirability bias and improve the quality of reporting; however, these methods often limit the complexity of information collected. We designed a life history calendar—the Relationship History Calendar (RHC)—to increase the scope of data collected on sexual relationships and behavior while enhancing their quality. The RHC records detailed, 10-year retrospective information on sexual relationship histories. The structure and interview procedure draw on qualitative techniques, which could reduce social desirability bias. We compare the quality of data collected with the RHC with a standard face-to-face survey instrument through a field experiment conducted among 1,275 youth in Kisumu, Kenya. The results suggest that the RHC reduces social desirability bias and improves reporting on multiple measures, including higher rates of abstinence among males and multiple recent sexual partnerships among females. The RHC fosters higher levels of rapport and respondent enjoyment, which appear to be the mechanisms through which social desirability bias is minimized. The RHC is an excellent alternative to private response methods and could potentially be adapted for large-scale surveys.     

A study of the correlation bias of unrecorded events by two independent enumeration systems

In this paper the allegation that, when using the Chandrasekar-Deming technique with two independent enumeration systems to record the incidence of vital events and migrations, each individual system misses the same type of event is investigated. If this correlation bias were to occur frequently enough, it could result in a serious underestimation of a given type of event. Using vital events and migrations data derived from a longitudinal survey conducted in Liberia between 1969 and 1973, the estimated number of recorded events is tabulated into homogeneous groups that are demonstrated to affect omission rates. The proportion of missed eyents is computed separately for each enumeration system, and the presence or absence of a significant correlation mathematically determined. The results of this study suggest that no significant correlation could be demonstrated in the type of birth, death or infant death that was commonly missed by each individual enumeration system. This indicates that the omission of a birth, death or infant death was apparently a random occurrence as far as the individual casefinding systems were concerned. The authors conclude that any underestimation of vital events from this type of correlation bias may not be as serious as originally imagined.

In contrast to this observation, the omission of particular types of migrations by each enumeration system was apparently not a random occurrence, but was associated with the direction of migration. This implies that selected migration rates in the liberian survey were probably understated and this could happen in other longitudinal surveys, unless specific precautions are taken to avoid it.     

Exposure to Violence,Coping Strategies,and Diagnosed Mental Health Problems Among Adults in a Migrant-Sending Community in Central Mexico

Deadly violence has spread throughout Mexico, affecting the well-being of citizens. What is the impact of this violence on the daily lives of Mexican adults? Building upon the stress process model, we used a mixed-methods approach to examine relationships between multiple indicators of exposure to and fear of violence and four diagnosed mental health outcomes, as well as self-rated mental health, in a Mexican community using the Survey of Health and Mexican Migration (456 surveys; 49 interviews). The multivariate models provide evidence that perceptions of insecurity are associated with diagnosed depressive episode, agoraphobia, alcohol abuse, a total count of mental health conditions, and poor self-rated mental health. Past victimization is associated with anxiety. Stress and coping behaviors did not formally mediate these violence–mental illness associations. The qualitative results confirm that residents fear violence and cope by adjusting their personal behaviors. These results foreshadow the emergence of mental health conditions as a critical public health concern for Mexicans living under the threat of violence.     

Comparing two survey methods for estimating maternal and perinatal mortality in rural Cambodia

PURPOSE: We need solid estimates of maternal mortality rates (MMR) to monitor the impact of maternal care programs. Cambodian health authorities and WHO report the MMR in Cambodia at 450 per 100,000 live births. The figure is drawn from surveys where information is obtained by interviewing respondents about the survival of all their adult sisters (sisterhood method). The estimate is statistically imprecise, 95% confidence intervals ranging from 260 to 620/100,000. The MMR estimate is also uncertain due to under-reporting; where 80-90% of women deliver at home maternal fatalities may go undetected especially where mortality is highest, in remote rural areas. The aim of this study was to attain more reliable MMR estimates by using survey methods other than the sisterhood method prior to an intervention targeting obstetric rural emergencies. PROCEDURES: The study was carried out in rural Northwestern Cambodia where access to health services is poor and poverty, endemic diseases, and land mines are endemic. Two survey methods were applied in two separate sectors: a community-based survey gathering data from public sources and a household survey gathering data direct from primary sources. FINDINGS: There was no statistically significant difference between the two survey results for maternal deaths, both types of survey reported mortality rates around the public figure. The household survey reported a significantly higher perinatal mortality rate as compared to the community-based survey, 8.6% versus 5.0%. Also the household survey gave qualitative data important for a better understanding of the many problems faced by mothers giving birth in the remote villages. There are detection failures in both surveys; the failure rate may be as high as 30-40%. PRINCIPLE CONCLUSION: Both survey methods are inaccurate, therefore inappropriate for evaluation of short-term changes of mortality rates. Surveys based on primary informants yield qualitative information about mothers' hardships important for the design of future maternal care interventions.     

A study of the correlation bias of unrecorded events by two independent enumeration systems

Summary In this paper the allegation that, when using the Chandrasekar-Deming technique with two independent enumeration systems to record the incidence of vital events and migrations, each individual system misses the same type of event is investigated. If this correlation bias were to occur frequently enough, it could result in a serious underestimation of a given type of event. Using vital events and migrations data derived from a longitudinal survey conducted in Liberia between 1969 and 1973, the estimated number of recorded events is tabulated into homogeneous groups that are demonstrated to affect omission rates. The proportion of missed eyents is computed separately for each enumeration system, and the presence or absence of a significant correlation mathematically determined. The results of this study suggest that no significant correlation could be demonstrated in the type of birth, death or infant death that was commonly missed by each individual enumeration system. This indicates that the omission of a birth, death or infant death was apparently a random occurrence as far as the individual casefinding systems were concerned. The authors conclude that any underestimation of vital events from this type of correlation bias may not be as serious as originally imagined. In contrast to this observation, the omission of particular types of migrations by each enumeration system was apparently not a random occurrence, but was associated with the direction of migration. This implies that selected migration rates in the liberian survey were probably understated and this could happen in other longitudinal surveys, unless specific precautions are taken to avoid it.     

Revising Infant Mortality Rates for the Early Twentieth Century United States

Accurate vital statistics are required to understand the evolution of racial disparities in infant health and the causes of rapid secular decline in infant mortality during the early twentieth century. Unfortunately, U.S. infant mortality rates prior to 1950 suffer from an upward bias stemming from a severe underregistration of births. At one extreme, African American births in southern states went unregistered at the rate of 15 % to 25 %. In this study, we construct improved estimates of births and infant mortality in the United States for 1915–1940 using recently released complete count decennial census microdata combined with the counts of infant deaths from published sources. We check the veracity of our estimates with a major birth registration study completed in conjunction with the 1940 decennial census and find that the largest adjustments occur in states with less-complete birth registration systems. An additional advantage of our census-based estimation method is the extension backward of the birth and infant mortality series for years prior to published estimates of registered births, enabling previously impossible comparisons and estimations. Finally, we show that underregistration can bias effect estimates even in a panel setting with specifications that include location fixed effects and place-specific linear time trends.     

Estimation of sample size of population survey]

Depending on the different purposes of population surveys, different sample sizes are required for accurate results. Using statistical methods, sample sizes for different types of population surveys for the Chinese population are estimated. Thus, to insure at least 90% accuracy, a sample size of 830,000 is required for marriage and birth rates and a sample size of 7.8 million for population age distribution. For population surveys on age specific death rates, a sample size of 56.32 million is required for a 1-year retrospective study and a sample size of 11.26 million required for a 5-year retrospective study. In the past, certain population studies were undertaken with no consideration for sampling errors. Sample sizes for surveys on age specific death rates and population age distribution were too small, leading to results that did not represent the total population. On the other hand, sample sizes for marriage and birth conditions were too large, resulting in waste in manpower, materials, and time. Statistical calculations are given.