you can find copy of scale with 1 to 5 ratings at:

http://breakingprejudice.org/assets/MaryKite/Homosexuality%20Attitude%20Scale.pdf

Thank you David L Morgan Sir for your kind response. I am aware of this PDF copy but unfortunately I found the scoring information limited. I can do the total score but not sure whether I can set the range by my own to level the attitude. (eg- High, Medium Low etc.)

It shows a Likert-score response format for each item that ranges from 1 to 5, Strongly Agree to Strongly Disagree. If you don't use a 1 to 5 format, you will not get a proper total score.

I understand David L Morgan Sir, my question is can I make range from the total score by my own since it is a standardized questionnaire. Thank you so much.

I am not sure if I understand your question. You should use the full range of the scale if you do correlations or t tests etc.

Yes David L Morgan Sir, the full range will be 21- 105. My question is can I prepare levels{ suppose High (21-50), Medium(51-70), Low( 71-105)} from within this full range? I hope now my question is clear. Thank you.

You should maintain the full range of scores in any analyses that you do. Collapsing them into 3 categories will produce a severe loss of information, and setting arbitrary boundaries for those categories is likely to get you in trouble with any reviewers who are familiar with this scale.

@David L Morgan Sir, since the questionnaire doesn't categorise the levels ( High, Medium and Low) and if I need those categories for analysis, in that case what should I do?

Can you say more about why you need that categorization for your analysis?

@ David L Morgan Sir, I want to distribute my respondents attitude based on those levels or categories. For example, I want to find out how many percentage from the total respondents have High/Medium/ Low level of attitude towards homosexuals. Thank you !

I am still unclear what you mean by "I want to distribute my respondents attitude based on those levels or categories." Is this simply for reporting purposes, such as a descriptive summary? Or do you what to create a 3-category independent or dependent variable? It is the second of these two options that you should not pursue, due to loss of information.

The outcome of HAS is continuous, from 21-105. And, there's no set criteria to divide the range into low medium and high attitudes. No matter what criteria you'll adopt will be criticised.

If you take mean instead of sum of the scale, you'll have mean between 1 to 5, which you may compare with the original measurement scale ( 1-5 point scale. ) This is not a foolproof approach, however, it'll give you cushion to explain mean score.