Bauhaus-Universität Weimar

Experimental Psychology: A Manual of Laboratory Practice, Vol. II: Quantitative Experiments, part 2: Instructor's Manual
Titchener, Edward B.
The Metrie Methods 
gradation (or the rate of continuous change) of rl is regulated by E\ O 
knows nothing of it. Moreover, the nature of the gradation is exactly 
recorded, so that the series can be repeated under precisely the same 
conditions. In our combined method, O adjusts rx for himself, and there 
is no guarantee that he can reproduce his adjustment in any later test. 
The psychological conditions of the judgment of disappearing difference 
are thus widely divergent in the two cases ; and with the difference in 
conditions goes a difference in psychological value. 
Again, our new method is, at the best, an incomplete application of 
the method of limits. For the method of limits determines not only the 
point of disappearing difference, Lut also the point of just noticeable 
difference. While, therefore, the method of limits can give us the values 
exl and el%—if those values are desired,—the combined method is unable 
to furnish the values Aru and Arh characteristic of the method of limits. 
The advantage is all on the side of the latter method ; the combined 
method has no raison d'être. 
' Not even as affording a measure of the j. u. d. ?’ No ! For we, do 
not really determine the rx that is j. u. or <(( r. Starting from rl |<(| 
r, we determine the lower limit of j. u. d. Starting from ^ |^>j r, we 
determine the upper limit of j. u. d. But these are not the upper and lower 
limits of one and the same j. u. d. The lower limit is the lower limit of 
an ascending j. u. d. ; the difference r — rx is referable, not to r, but to 
rv The upper limit is the upper limit of a descending j. u. d. ; the 
difference rt—r is referable, not to rlt but to r. We ought by rights, in 
our ^ adjustments, to determine the first point of equality and the last 
point of equality of r and ru and then to strike the average. Similarly, 
we ought, in our ^J, adjustments, to determine both the first and the last 
points of equality, and to average. As the j. n. d. is the mean between 
first difference and disappearing difference, so is the j. u. d. the mean 
between disappearing difference and last equality. Why, then, do we 
not seek to determine this mean ? Simply because the determination is 
impracticable. We can, without too great difficulty, arrest the moving 
screen at the point where rl first seems equal to r. But if we try to find 
the point at which rl last seems equal to r, we either stop timidly at 
some point within the zone of equality, or we overshoot the mark and 
stop only at some point of positive difference. 
This latter criticism may, perhaps, be discounted in practice. The 
twoj. u. d. will, very probably, overlap ; 'and the averaging of the 'J' and 
^ values would not lead to any considerable error. Still, the theoretical 
error remains. And in view of the imperfections of the method, taken 
as a whole, we may surely say, without hesitation, that it is not worth pre¬ 
serving. If one is forced for any reason to work by oneself, the combined


Sehr geehrte Benutzer,

aufgrund der aktuellen Entwicklungen in der Webtechnologie, die im Goobi viewer verwendet wird, unterstützt die Software den von Ihnen verwendeten Browser nicht mehr.

Bitte benutzen Sie einen der folgenden Browser, um diese Seite korrekt darstellen zu können.

Vielen Dank für Ihr Verständnis.