Ratio Variables in Aggregate Data Analysis

Bollen, Kenneth A.; Ward, Sally

doi:10.1177/004912417900700405

Cited by 50 publications

(5 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For this reason, the actual counts of the three types of incidents that occur within each area are retained as measures. To control for the effects of area size and population, these measures are introduced as independent variables in the multiple regression analysis (Agresti and Finlay 1997;Bollen and Ward 1980). 16 …”

mentioning

confidence: 99%

The Impact of Neighborhoods, Schools, and Malls on the Spatial Distribution of Property Damage

Lagrange¹

1999

Journal of Research in Crime and Delinquency

View full text Add to dashboard Cite

Using data obtained from three different sources, principles derived from routine activities theory are used to predict the distribution of minor property crimes in a medium-sized Canadian city during a 1-year period. Mischief and vandalism incidents recorded by the local police, transit department, and department of parks and recreation are aggregated by census enumeration area using mapping software, and analyzed in relation to three sets of predictors: (1) neighborhood demographic characteristics; (2) the proximity of shopping malls; and (3) the proximity of public and Catholic senior and junior high schools. Similar patterns observed for the three types of damage are examined in relation to the convergence of potential offenders, reduced guardianship, and increased opportunity as derived from routine activities theory.

show abstract

mentioning

confidence: 99%

The Impact of Neighborhoods, Schools, and Malls on the Spatial Distribution of Property Damage

Lagrange¹

1999

Journal of Research in Crime and Delinquency

View full text Add to dashboard Cite

show abstract

“…The key problem for aggregate data is that, because the data are aggregated, one cannot reasonably infer the effects of extraneous information or the importance of the information that is lost during aggregation. This point has been made by a number of scholars, including Bollen and Ward (1979), Fox (1984), and Hanushek and Rivkin (2006). Because of this, wide deviations at one level can be rolled up to the next level, and one who sees the final aggregate product will have no knowledge of these deviations.…”

Section: The Issue Of Lists and Public Informationmentioning

confidence: 93%

Changing Ethics Policies Without Scandal: State Responses to Published Reports and the Importance of Accurate Information

Rauh

2016

Public Integrity

View full text Add to dashboard Cite

This article details how states responded to the State Integrity Investigation-a large-scale study that sought to rank states based on their risk of corruption. Although some states responded with policy changes, the article demonstrates that responses were not due to how well the states performed compared to other states. Rather, their ranks on the survey magnified the effect of prior information on the state's ethics policies. Additionally, it goes on to show flaws in the ranking methods used in the State Integrity Investigation and demonstrates that these ranks likely created inconsistent responses from states with regard to changing ethics policies.

show abstract

“…Then, we performed a series of multivariate longitudinal analyses of GRP on RTFs and CFR, adjusting for population and other socioeconomic variables (Bollen and Ward 1979; Schuessl 1974). …”

Section: Methodsmentioning

confidence: 99%

“…Both road safety and socioeconomic variables were entered in the model in their aggregate values. We controlled for population size by including population as a covariate in the regression model to avoid the well-known ratio bias that comes from dividing both dependent and independent variables by population size (Bollen and Ward 1979; Schuessl 1974). Both fixed effects and random effects models were fitted to compare their goodness-of-fit to the panel data for federal regions.…”

Section: Methodsmentioning

confidence: 99%

Economic development and road traffic fatalities in Russia: analysis of federal regions 2004–2011

Paichadze

Hyder

et al. 2015

Inj. Epidemiol.

View full text Add to dashboard Cite

BackgroundThe relationship between economic development and road safety at sub-national level has not been well established. This study aims to assess the relationships between economic growth (measured by gross regional product (GRP)) and road traffic fatalities (RTFs) and crash fatality ratio (CFR) at sub-national level in Russia.MethodsWe used published secondary data on annual RTFs and CFR obtained from the traffic police and socioeconomic development indicators from the statistics department for each Russian federal region (referred to in Russia as “subject”) for 2004–2011. We used multivariate fixed effects models for longitudinal data to examine the GRP-RTF and the GRP-CFR relationships excluding regions with extreme values. Time (in years) and a set of relevant socioeconomic variables (territory, population, number of privately owned cars, number of public buses, length of public motor roads, number of physicians, and budget expenditure on health care and physical wellness) were also included as covariates in the models.ResultsThe RTF rates decreased monotonically over time as GRP per capita increased in 66 studied regions during 2004–2011. This relationship was mainly explained by the number of privately owned cars and partially explained by year dummy variables, number of buses, and number of physicians. CFR also decreased monotonically as GRP per capita increased in 67 studied regions. This relationship between economic growth and CFR was fully explained by secular time trends. The year dummy effects on CFR were not mediated by other socioeconomic variables included in the study.ConclusionsFor the period of 2004–2011 in Russia, the reduction in RTFs is mostly explained by increasing the number of private cars, while the reduction of CFR is mostly associated with year-effects suggesting a process of diffusion of knowledge, which is not solely dominated by economic growth.

show abstract

Ratio Variables in Aggregate Data Analysis

Cited by 50 publications

References 32 publications

The Impact of Neighborhoods, Schools, and Malls on the Spatial Distribution of Property Damage

The Impact of Neighborhoods, Schools, and Malls on the Spatial Distribution of Property Damage

Changing Ethics Policies Without Scandal: State Responses to Published Reports and the Importance of Accurate Information

Economic development and road traffic fatalities in Russia: analysis of federal regions 2004–2011

Contact Info

Product

Resources

About