Motels and Hotels
 A total of four hotel-motel models represented by Equations
 (3.1-3.4) in Table 5 were estimated for the motel-hotels group.
 Equations (3.1) and (3.2) were fit to the primary data. Equa-
 tions (3.3) and (3.4) are based on the secondary data. Hetero-
 skedasticity problems were found in both the primary and
 secondary data. The models were adjusted to reduce the effects
 of this problem.'1 Nearly all coefficients in the adjusted models
 were significant at (at least) the 0.01 level (Table 5). The area
 of dining room plus bar room area (DB variable) was found to
 be not significant. The R' estimates were high at 0.95 and 0.94
 for the primary and secondary models, respectively. The mean
 quantities for each of the variables are also reported in Table 5.
 The relationship between the quantity purchased and the in-
dependent variables was generally as expected for this group.
Quantity purchased per month was shown to increase with the
size of the establishment (measured by number of rooms), the
area of the restaurant and lounge area, and the price of the
rooms. Also, as expected, the quantity purchased per month de-
clined with increases in the price of water.
 Differences in estimates from the two models (Equations 3.2
and 3.4) were not great with the exception of very large motels
or hotels, as shown in Figure 3. For hotels having less than 54
rooms, both demand curves predicted approximately the same
quantities of water. At a price of $0.40 per thousand (and a
motel size of 54 rooms), the primary model predicted a monthly
consumption of 193,000 gallons, with the secondary motel pre-
dicting 153,000 gallons. For higher prices, the differences are
even less. The difference was only 3,000 gallons at a price of
$2.00 per thousand gallons.
 The size of establishment variable affected purchases in a
different manner than was the case for department and grocery
stores. Given any particular price, water purchases were shown
to more than double for a doubling in the size of the motel-
hotel. For example, at a price of $0.40 the purchases were 193,-
000 gallons for the 54 room unit and 846,000 gallons for the 108
room unit. This suggests the quantity of water purchased per
month is elastic with respect to a change in the number of
rooms. This elastic relationship holds for all establishments
having 36 or more rooms, otherwise the relationship is inelastic.
 Price elasticity varied between the primary and secondary
data models. Both models revealed inelastic demand over the
 12See Appendix A.