More ratio stuff

[From Bruce Abbott (950802.2120 EST)]

This is getting to be a regular "feature." I've taken a look at some more
of the data Ettinger, Thompson, & Staddon (1986) present. Last time I
discussed data in which the "bar weight" was varied. The same animals were
also tested following injection of a drug, cholecystokinin (CCK-8), which,
according to the authors, may act "as a short-term satiety signal. As
before, I have measured the food rates (x-intercept) and response rates
(y-intercepts) from the figure and then converted these into "collection
times" and "lever-press times" (in seconds). Here are the results,
including the saline figures previously posted as "45 gm bar":

Rates (per minute):

Rat drug y-intercept x-intercept
             (resp/min) (pellets/min)
CK1 saline 130 10.3
      CCK-8 95.2 6.5

CK2 saline 103 10.9
      CCK-8 82.2 8.9

CK3 saline 71.7 10.7
      CCK-8 43.2 4.7

CK4 saline 125 10.7
      CCK-8 80.3 4.2

Times (seconds):

Rat drug collection seconds/
                  time response
CK1: saline 5.8 s 0.46 s
       CCK-8 9.2 s 0.63 s

CK2: saline 5.5 s 0.58 s
       CCK-8 6.8 s 0.73 s

CK3: saline 5.6 s 0.84 s
       CCK-8 12.7 s 1.39 s

CK4: saline 5.6 s 0.48 s
       CCK-8 14.4 s 0.75 s

As seen on the original figure (depicting response rate as a function of
food rate), the lines representing saline and CCK-8 are roughly parallel,
with the CCK-8 functions displaced to the left of the saline functions. The
lines appear parallel for rat CK2; for the remaining rats the lines tend to
diverge slightly toward the bottom. The authors note that the slopes of the
saline and CCK-8 functions are not significantly different (statistically

A second group of rats received either saline or a low-dose lithium chloride
injection, with similar results. (I won't present the data here.) Lithium
chloride at higher doses tends to make the rats ill.

The authors believed that the functions would allow them to separate the
effect of these drugs "degree of food-regulation" (gain, supposedly
indicated by the slope of the response function in the original figure) and
"incentive" factors such as palatability. Factors which affect the rat's
willingness to defend the set-point for food rate are said to alter the
slope, whereas those which affect incentive factors alter the x-intercept
without affecting the slope. They concluded that the drugs appear to change
incentive without affecting regulation (gain).

What is actually happening is that the drugs increase both the time per
lever-press and the time to collect the pellet, in the same way that
deprivation does. There is no way from these data to determine whether the
suppression is due to reduction in palatability of the food, increased
perceived effort to obtain the food, reduced set-point in the
nutrient-control system (i.e., reduced appetite), or some combination of the
above, and certainly no way to assess any effect on "regulation." All we
learn is that these drugs reduce the rates at which the rats are willing to
work in order to obtain the food, by somewhat different amounts for
different rats.