“Halo Effect: A Constant Error in Psychological Ratings”, 1920 (; backlinks):
In a study made in 1915 of employees of two large industrial corporations, it appeared that the estimates of the same man in a number of different traits such as intelligence, industry, technical skill, reliability, etc., etc, were very highly correlated and very evenly correlated. It consequently appeared probable that those giving the ratings were unable to analyze out these different aspects of the person’s nature and achievement and rate each in independence of the others Their ratings were apparently affected by a marked tendency to think of the person in general as rather good or rather inferior and to color the judgments of the qualities by this general feeling This same constant error toward suffusing ratings of special features with a halo belonging to the individual as a whole appeared in the ratings of officers made by their superiors in the army.
…Mr. Knight of Teachers College has studied this same effect in the case of 129 teachers rated by their superior officer for certain qualities on the Boyce score card. The ratings in question were official and were used to determine salaries and promotions. General merit as a teacher has correlations of 0.68 with intellect, 0.79 with power in discipline, and 0.63 with voice. It is clear that the rating of a teacher’s voice must have been influenced by the general impression of her ability. Voice correlates 0.50 with “Interest in Community Affairs”, and 0.63 with intelligence!
…In the cases so far the correlations are a resultant of (1) the real facts, (2) the constant error of the “halo”, as we may call it, and (3) the reverse error of [measurement-error] attenuation due to chance inaccuracies in the ratings. In certain further work by Mr. Knight the correlations are freed from the last influence, by being based on the composite rating of two groups, each of a number of teachers who knew the individuals to be rated fairly well. The self-correlations of the ratings by one such a group with ratings for the same trait by the other group are over 0.90. The correlations for general ability as a teacher with intellect and with ability to discipline are about 0.95 and 0.80! The correlation of intelligence and ability to discipline is about 0.80! The correlations of a standard test of intelligence with general ability as a teacher and with ability to discipline are, for the individuals in question, not over 0.3.
The writer has become convinced that even a very capable foreman, employer, teacher, or department head is unable to treat an individual as a compound of separate qualities and to assign a magnitude to each of these in independence of the others.
The magnitude of the constant error of the halo, as we have called it, also seems surprisingly large, though we lack objective criteria by which to determine its exact size.
See Also: