Efficient and effective learning for safety from incidents

Learning from incidents is important for improving safety. Many companies spend a great deal of time and money on such learning procedures. The objectives of this paper are to present some early results from a project aimed at revealing weaknesses in the procedures for learning from incidents and to discuss improvements in these procedures, especially in chemical process industries. The empirical base comes from a project assessing organizational learning and the effectiveness of the different steps of the learning cycle for safety and studying relations between safety-specific transformational leadership, safety climate, trust, safety-related behavior and learning from incidents. The results point at common weaknesses in the organizational learning, both in the horizontal learning (geographical spread) and in vertical learning (double-loop learning). Furthermore, the effectiveness in the different steps of the learning cycle is low due to insufficient information in incident reports, very shallow analyses of reports, decisions that focus at solving the problem only at the place where the incident took place, late implementations and weak solutions. Strong correlations with learning from incidents were found for all safety climate variables as well as for safety-related behaviors and trust. The relationships were very strong for trust, safety knowledge, safety participation and safety compliance.     

Measurement errors in survey questions: Explaining response variability

In this paper, the response variability in questions from the Swedish Survey of Living Conditions, is studied. A multivariate study of the variation over questions has been done. This study uses questions as statistical units. The questions have been classified in a number of respects, dealing with question character and structure, what the questions ask about and what they require from the interviewee. The study shows that a substantial part of the response variability, i.e., the variation in reliability, can be explained by a few characteristics of the questions.     

High-risk families: the unequal distribution of infant mortality in nineteenth-century Sweden

An analysis of infant mortality (based on 133,448 births) in two regions, Sundsvall and Skellefte?, in north-eastern Sweden during the nineteenth century shows that infant mortality was highly clustered with a relatively small number of families accounting for a large proportion of all infant deaths. Using logistic regression, two important factors were found to be associated with high-risk families: (i) a biological component evidenced by an over-representation of women who had experienced stillbirths, and (ii) a social component indicated by an increased risk among women who had remarried. The results strengthen the argument for using the family rather than the single child as the unit of analysis. The clustering of infant deaths points to the need to re-evaluate our interpretations of the causes of infant mortality in the past.     

Demographic implications of the New United States certificates

To assist in developing uniform reporting of vital events among the fifty states, Puerto Rico, and the Virgin Islands, the United States government prepares standard certificates of birth, fetal death, death, marriage, and divorce. These model forms are revised, with the assistance of the states, approximately every ten years. Revisions are now being prepared by the National Center for Health Statistics which will become effective beginning January 1, 1968. Important new source material for demography will be introduced.Most changes will appear in the Standard Certificate of Live Birth and in the Standard Certificate of Fetal Death. An item on education of father and mother will provide detailed national data on education and fertility. The date of the last live birth to the mother and the date of the last fetal death will provide information on previous pregnancy outcome and on child-spacing. The recording of state file numbers for mates born alive and dead in the same delivery will make it easier to match live birth and fetal death certificates for the preparation of detailed tabulations on multiple births. Several new items related to maternal and child health have also been added. No significent changes were planned for the Standard Certificate of Death.The Standard Certificate of Marriage will include as new items the education of the bride and groom, the date on which the last marriage, if any, ended, and specification of the officiant as a religious or civil official. The Standard Certificate of Divorce or Annulment will obtain information on the education of husband and wife, the approximate date on which the couple separated, the mode of dissolution of the previous marriage, and the total number of living children. It is anticipated that most of the new items will be included in the certificates of all the states. The National Center for Health Statistics will provide detailed tabulations related to these items, beginning with data year 1968.Demographers are making an increased use of vital records and at the same time are extending their contacts with state health departments; in some states collaborative projects have been undertaken. Because of the importance of the source documents, which in some cases have not been exploited fully, demographers should increase their contact with the state vital statistics offices which develop, collect, and process the records. By indicating an interest in vital registration and by making their research needs known, demographers can encourage the acceptance of new concepts and collaborate in the improvement of vital records for demographic research purposes.     

Integration of Childhood Immigrants in the Short and Long Run – Swedish Evidence1

Immigrant children are often at a disadvantage at school while they adjust to their new environment. It has been found that the age at immigration of 7 or above represents a sensitive period as regards these children’s school performance for two main reasons: they have passed their prime age for language learning, and their acquisition of subject skills is less efficient while they learn to master the new language. Using Swedish administrative data, we track childhood immigrants born between 1972 and 1976 over time from adolescence (at 16) to adulthood (at 30), and study the role of age at immigration for educational and labor market outcomes. We find that immigration at a sensitive age (compared to a very young age) has a strong negative impact on compulsory school performance, but that the same individuals catch up strongly in terms of final educational attainment. In spite of this educational catching up, however, we find a considerable negative impact on earnings for men. We are able to rule out two potential mechanisms behind this puzzling result: It cannot be explained either by an impact on the chosen field of study or on completion time.     

Bootstrap Algorithms for Testing and Determining the Cointegration Rank in VAR Models1

In this paper a bootstrap algorithm for a reduced rank vector autoregressive model with a restricted linear trend and independent, identically distributed errors is analyzed. For testing the cointegration rank, the asymptotic distribution under the hypothesis is the same as for the usual likelihood ratio test, so that the bootstrap is consistent. It is furthermore shown that a bootstrap procedure for determining the rank is asymptotically consistent in the sense that the probability of choosing the rank smaller than the true one converges to zero.     

Statistical Corrections of Invalid Correlation Matrices

Suppose estimates are available for correlations between pairs of variables but that the matrix of correlation estimates is not positive definite. In various applications, having a valid correlation matrix is important in connection with follow‐up analyses that might, for example, involve sampling from a valid distribution. We present new methods for adjusting the initial estimates to form a proper, that is, nonnegative definite, correlation matrix. These are based on constructing certain pseudo‐likelihood functions, formed by multiplying together exact or approximate likelihood contributions associated with the individual correlations. Such pseudo‐likelihoods may then be maximized over the range of proper correlation matrices. They may also be utilized to form pseudo‐posterior distributions for the unknown correlation matrix, by factoring in relevant prior information for the separate correlations. We illustrate our methods on two examples from a financial time series and genomic pathway analysis.     

Incomplete equalization: The effect of tracking in secondary education on educational inequality

This paper tests whether the existence of vocationally oriented tracks within a traditionally academically oriented upper education system reduces socioeconomic inequalities in educational attainment. Based on a statistical model of educational transitions and data on two entire cohorts of Danish youth, we find that (1) the vocationally oriented tracks are less socially selective than the traditional academic track; (2) attending the vocationally oriented tracks has a negative effect on the likelihood of enrolling in higher education; and (3) in the aggregate the vocationally oriented tracks improve access to lower-tier higher education for low-SES students. These findings point to an interesting paradox in that tracking has adverse effects at the micro-level but equalizes educational opportunities at the macro-level. We also discuss whether similar mechanisms might exist in other educational systems.