APPENDIX II 

 Scalogram Analysis 



The principal of scalogram analysis is one of ranking people from more 

 favorable to less favorable in their attitude toward something. The ranking 

 is made bv their responses to a series of questions, each response being con- 

 sidered as a separate item. The rank order must have a special cumulative 

 property such that all persons who answer a given question favorablv must 

 have higher ranks than persons who answer the same question unfavorably. 

 Such a ranking is called a perfect scale because it has the property of per- 

 fect internal consistency. 



Assuming that the universe of content I the subject which is being measur- 

 ed I is made up of a series of questions favorable in nature, the above defini- 

 tion of a perfect scale leads to a parallelogram response pattern. Consider a 

 hypothetical example of three questions from a content universe. A. B. and 

 C. each question graded in degree of favorableness from more favorable to 

 less favorable with the response categories dichotomized into "yes" (Aj, Bj. 

 Cj I and "no" (A 2 , B 2 . C 2 ) answers: 



Rank Order Type of Category i Item Response > 



Respondent of Replies "yes Replies "no* 



Number Respondents to question to question 



Ai Bj Ci A 2 B-2 Co 



1 4 



2 3 



3 2 



4 1 



XXX 



XXX 



X X \ 



X X 



If a scale exists according to the above definition, there are only four pos- 

 sible response patterns. A person who checks \ 1 must also check B r and 

 C 1 : a person who checks B x must also check C x and Ao : and so forth. The 

 resulting parellelogram arrangement is called a scalogram and ranks the 

 population from more favorable to less favorable toward something. This 

 rank order is a linear function of the scores assigned to each item. Thus, if 

 the population rank order is related to the scores assigned to the answer 

 responses, the higher the rank the higher will be the score as shown in the 

 chart at the top of page 28. 



The relationship between the population rank and the content score is 

 the first component of an attitude. The first component does not determine 

 if the population under study is favorable or unfavorable to the content 

 area, but only that they rank from more favorable to less favorable. 



The technique does allow, however, a determination of the proportion of 

 the population favorable or unfavorable by an intensity analvsis. Intensity 

 is the second component of an attitude. Bv relating the rank order of content 



27 



