Academic journal article North American Journal of Psychology

Substitution of Method in Suicide and Homicide: An Ecological Analysis

Academic journal article North American Journal of Psychology

Substitution of Method in Suicide and Homicide: An Ecological Analysis

Article excerpt

I examined crude and age-adjusted rates of homicide and suicide for 1980 and 1990 for the 48 continental states. Methods of homicide were statistically associated with each other and with the total homicide rate. For 1980 and 1990 rates of suicide by poisons, hanging/strangulation, and other methods all correlated positively. Further, in 1980 all three of these rates correlated with the suicide rate by firearms, whereas in 1990 the latter two rates were negatively correlated. Of these two the only significant correlation was for the 1990 rates of suicide by firearms and other methods. There seems to be a tendency for persons in different regions to use one method or another to commit suicide. Generally speaking, the more firearms and poisons are used to commit suicides, the less hanging and other methods are used.

Lester (1988) reported in a time series that rates of homicide by each of several methods positively correlated with each other in the UK and USA. In the same study Lester reported that rates of suicide by hanging and all other methods negatively correlated with rates of suicide using firearms and poisoning. He noted that as one method of suicide became less popular, other methods of suicide became more popular. Drawing upon these time series results for homicide and suicide Lester (1987) conducted an ecological study across the 48 contiguous states in the United States. The purpose of the present study was to determine if Lester's (1987) ecological correlations from 1970 data produced a similar pattern in 1980 and 1990 correlations.

Crude and age-adjusted rates (1) for each method of homicide and suicide were obtained for 1980 and 1990 for each of the 48 contiguous United States from data available at WISQARS (2) and CDC WONDER (3). WISQARS contained only data back to 1981; however, CDC WONDER had 1980 data. WISQARS figures the crude rate per 100,000 by dividing the number of deaths in a particular population by the total number of people in that population, then multiplying that ratio by 100,000. WISQARS calculates age-adjusted rates by the direct method and standardizes them to the total U.S. population. The standard population used in this study was the year 1990. On the other hand, CDC WONDER figured age-adjusted rates by calculating with non-standard population denominators using the year 1980.

For 1980, the three rates of homicide (by firearms and explosives, cutting and piercing, and other means) were significantly associated with each other (rs = .76, .63, .78) in Table 1. However, because the likelihood of Type I error is quite large with the number of correlations in the present study the author sought stronger evidence of significant associations beyond what would have been expected by chance. Thus, the reportable alpha was adjusted from .05 to .01 or .001 to control for that likelihood. All three rates of homicide (by firearms and explosives, cutting and piercing, and other means) correlated significantly with the total homicide rate (rs = .98, .86, .76) and again all three correlations were positive.

In 1990, the same pattern was revealed. The three rates of homicide (by firearms and explosives, cutting and piercing, and other means) were positively and significantly correlated (rs = .87, .70, .81). Each of these three rates for homicide in 1990 correlated positively and significantly with the total rate of homicide (rs = .99, .93, .80). These homicide findings for 1980 and 1990 match those reported by Lester (1987) using data for 1970.

The 1980 rates for suicide by poisons, hanging/strangulation/suffocation, and other methods all correlated positively (rs = .43, .48, .46) and these findings presented in Table 2 were in agreement with those of Lester (1987). Unlike Lester (1987), all three (not just two) of these rates of suicide correlated negatively with the rate of suicide by firearms (rs = -.05, -.43, -.49, respectively). Two out of these three correlations were significant. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.