Question

Stevenson et al (1999) conducted an analysis of violence in suburbs of NSW, using data for 178 suburbs on:

ASSAULTS: assaults that took place in each suburb (“the number of assaults reported to police between January and December 1995”)

ALC_SALES: quantity, in litres, of pure alcohol sold in each suburb between January and December 1995. (“sales of alcohol from hotels, clubs, restaurants and off-licenses”) Consider the following illustrative EViews output:

Dependent Variable: ASSAULTS                 
Method: Least Squares                    
Sample: 1 178                  
Included observations: 178

Variable Coefficient Std. Error   t-Statistic    Prob.      
C         16.86111    2.109365     7.993454      0.0000
ALC_SALES 0.062804    0.026214     2.395819      0.0179

For a suburb with alcohol sales of 1,000 litres during the period, predict how many assaults took place.

Answer 1

Hello

Using the given data, following relevant vlues can be extracted:

Variables	Coefficients
C	16.86111
ALC_SALES	0.062804

Hence, using this data, we can easily write the regression equation:

y = (0.062804*x) + 16.86111

where y = Number of Assaults and x = Sale of Alcohol, in litres.

Hence, to find the number of assaults if the alcohol sale is 1000 litres, put x = 1000

then,

y = (0.062804 * 1000) + 16.86111 = 62.804 + 16.86111 = 79.66511 assaults 80 assaults.

I hope this solves your doubt.

Do give a thumbs up if you find this helpful.