Page 97 - 2022 Spring JSOM

negative relationship with Children with Educational Needs to to the Coefficients column in Table 3) for Persons with Pri-
the dependent variable, terrorist attacks. mary Health Needs, is 7.012. Interpreting the results of a log
linear regression model is not always perfectly straightforward
Although we must not dismiss the value of such humanitarian because most people do not think in logarithmic terms. One
necessities as water, sanitation, and hygiene (WASH), shelter, common way to interpret these coefficients would be to say, in
or rehabilitation needs, these simply do not appear to sub- the case of Persons with Primary Health Needs, that a 1% in-
stantially affect the outcome we are seeking to influence; thus, crease in the number of people with primary healthcare needs
expending scarce resources pursuing them may prove to be is associated with a 7% increase in the incidence of terrorist
inefficient, wasteful, and even detrimental to the overall goal attacks in any given governate. Conversely, a 1% decrease in
of lowering the incidence of terrorist attacks. Furthermore, the number of people with primary healthcare needs is associ-
according to this model, negative coefficients are present for ated with a 7% decrease in the incidence of terrorist attacks in
internally displaced persons, WASH, shelter, and educational any given governate.
needs, which implies an interesting relationship to terrorist at-
tacks once the logarithmic manipulation of the data is taken Frankly speaking, so what? Indeed, while statistics can yield
into account. The negative coefficients show that as these interesting correlations, this by no means implies causality
specific public health needs worsen, the incidence of terrorist and still requires thoughtful analysis to understand if the re-
attacks moves exponentially closer to zero because the coeffi- sults make sense. From an operational perspective, it would be
cient values are negative. Table 4 captures these findings in the necessary to contextualize the findings within an established
same format that the linear regression results are published in doctrinal construct such as PEMESII-PT (Political, Military,
Microsoft Excel. Economic, Social, Infrastructure, Information, Physical Envi-
ronment, and Time) to ensure that the results are indeed rea-
sonable. Given that the implication of our findings is that a
Discussion
public health need correlates to terrorist attack incidence, then
Assuming there are opportunity costs for pursuing a specific a public health framework, such as the social determinants of a
line of effort, what is the anticipated return on investing time health model, may prove to be a useful approach for analysis.
and effort into improving primary health needs, should this In this context, economic stability, education, social cohesion,
be pursued? Inevitably, there will be limitations in resources, access to and quality of healthcare, and the built environment
such as money or time; therefore, addressing this independent all contribute to the health outcomes of individuals. 7
variable should have the potential to exert a negative effect on
the dependent variable while facilitating a need among the af- Knowing that most of these attacks were attributed to the
fected population. Examining the data output, the coefficient Islamic State, it is reasonable to assume that the young men
(i.e., the slope, m, from Line Equation 1, which corresponds who compose its rank and file are engaged in violence that
TABLE 4 Microsoft Excel Linear Regression Results
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.948945501
R Square 0.900497563
Adjusted R Square 0.830845857
Standard Error 0.673918235
Observations 18
ANOVA
df SS MS F Significance F
Regression 7 41.10202707 5.871718152 12.92857874 0.000276577
Residual 10 4.541657881 0.454165788
Total 17 45.64368495
Standard
Coefficients Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept -10.62417817 8.448785104 -1.257479986 0.237152753 -29.44924451 8.200888172 -29.44924451 8.200888172
X1 (Persons in Need) 9.037951407 5.21888554 1.731778047 0.113987563 -2.590450229 20.66635304 -2.590450229 20.66635304
X2 (Internally -0.026775609 0.470980391 -0.056850793 0.955783958 -1.076185317 1.022634099 -1.076185317 1.022634099
Displaced Persons)
X3 (Persons with 7.011873121 2.716720563 2.581006385 0.027369013 0.958642484 13.06510376 0.958642484 13.06510376
Primary Health Needs)
X4 (Persons with
Physical or Mental 3.654395688 1.975627922 1.849738833 0.09408539 -0.747577643 8.056369019 -0.747577643 8.056369019
Rehab Needs)
X5 (Persons with -0.207494727 0.266269296 -0.779266442 0.453869423 -0.800779691 0.385790237 -0.800779691 0.385790237
WASH Needs)
X6 (Persons with
Shelter and Non-Food -0.476867303 0.217939569 -2.188071238 0.053510216 -0.962466924 0.008732318 -0.962466924 0.008732318
Needs)
X7 (Children with
Educational Needs) -18.12993896 6.729401767 -2.694138289 0.022539018 -33.12398048 -3.135897431 -33.12398048 -3.135897431
df, degrees of freedom; F, distribution; MS, mean square; SS, sum of squares.

Prioritization of Humanitarian Efforts | 95