What problems might arise in trying to measure attitudes?
What problems might arise in trying to measure attitudes?
The question what problems might arise in trying to measure attitudes itself conjures up many complicated questions that seem tricky to answer. The complicated task is that we are all unique and have our individual differences.
A basic assumption on measuring attitudes is that 'a person's attitude can be measured by asking questions about thoughts, feelings and likely actions toward the attitude object
(Hoff and Vaughan, 1995, Chapter 5)
From this assumption it is obvious that measuring attitudes is not directly observable. Even if we think we know someone's attitude this does not mean we can predict their behaviour.
Measuring attitudes plays a large part in social psychology. Techniques have been developed in order to measure attitudes. "attitudes can be measured by a quantitative technique for example each persons opinion can be represented by a numerical scale"
(Hoff and Vaughan, 1995, Chapter 5.)
One of the initiators of attitude measurement was L.L Thurstone, he termed attitude measurement as "attitude is the intensity of positive or negative affect for or against a psychological object" (1946, page 39, http://webspace.utexas.edu/brixey9/www/frame.html.)
This section of attitude measurement is extremely vast and there is many aspects such as self reported test, indirect and multiple-indictor tests, likert scales, thurstones scale of attitude measurement and semantic differentials.
To measure the cognitive component of attitudes rating scales have been developed. The first mayor and a popular technique off attitude measurement was that of L.L Thurstone (1928), L.L Thurstone made this distinction between attitudes and opinions, 'he adds and attitude is "the sum total of a mans inclinations and feelings prejudice or bias, preconceived notions, ideas, fears, threats and convictions about any specified topic." He states a opinion as a "verbal expression of attitude." (http://webspace.utexas.edu/brixey9/www/frame.html.)
He theorised that 'the best way to measure a persons attitude was to evaluate their opinions. (http://webspace.utexas.edu/brixey9/www/frame.html.)
Thurston, to obtain a person attitude got together 130 statement he presenting favourable and unfavourable attitudes, a large group of judges sorted these statement into categories labelled A to K, F was a neural position. Next twenty of the statements that had the greatest agreement form the judges made the final scale, each was given a value arrived at by averaging all of values,
(Birch, A & Malim, T, 1998:659)
the judges ratings. Next the participants were presented with the statements in a random order and asked to indicate those with which they agreed. The numerical values attached to these statements were then averaged and a measure of participants attitude was measured and obtained. (Birch, A & Malim, T, 1998:659)
The problem that may arise here is in measuring attitudes is that the group of judges may be biased in their own opinions and views. Also problems could arise from the participants, the method is ambiguous. Thurston wanted the method ...
This is a preview of the whole essay
(Birch, A & Malim, T, 1998:659)
the judges ratings. Next the participants were presented with the statements in a random order and asked to indicate those with which they agreed. The numerical values attached to these statements were then averaged and a measure of participants attitude was measured and obtained. (Birch, A & Malim, T, 1998:659)
The problem that may arise here is in measuring attitudes is that the group of judges may be biased in their own opinions and views. Also problems could arise from the participants, the method is ambiguous. Thurston wanted the method to be objective through the sorting process, this has been largely criticized by Hovland and Sheriff _1952) Biases were found black judges or people that are sympathetic to black people sorted statements as compared to those who were unsympathetic to black people.
(Birch, A & Malim, T, 1998:659)
The level of measurement of the scale was also criticised, as Thurstone thought that the scale of measurement was interval scale date, the problem that the judges can not be completely objective meant the scale was ordinal. Thurstones method was self reported attitude measurement.
Renesis Likert (1932) followed from thurstones work Likert held the opinion that attitudes could be measured. In Likerts further study form Thurston s work he discovered that creating the initial attitude scale, problems appeared. The main problem was that
'the scale construction was that the statements selected by the researchers were dependant on the readers judgement and therefore biased.
(24/11/2003, Attitude Measurement, http://webspace.utexas.eduedu/brizey9/www/4.html, page2of 3)
Likert continued in his criticism and said 'the entire process of creating an attitude scale was 'exceedingly laborious'.
(24/11/2003, Attitude Measurement, http://webspace.utexas.eduedu/brizey9/www/4.html)
Likert constructed a newer attitude measurement instead of judges, respondents place themselves on an attitude continuum, that is, being given a question and having a scale of answers to pick from for example the answers would be;
. Strongly agree 2.Agree 3.Unsure 4.disagree 5. Strongly disagree.
The numbers on the side were not a indicator of what was presented to the participants, they would be used to attach a numerical value to responses, from the example of 5. Strongly agree etc, the highest numerical value represents the strongest unfavourable attitude. .
(Birch, A & Malim, T, 1998:659)
Likert quoted the following in order to certify/show that this method was appropriate to measure attitudes;
. The statement was an expression of ;desired behaviour' rather than fact.
2. Statements should be clear, simple and straightforward, avoiding any kind of ambiguity.
3. Questions should be designed so that responses lean to one side of the continuum.
4. Statements should be scattered randomly.
5. The alternatives in multiple choice questions should involve only on attitude variable
(Likert 1993) (24/11/2003, Attitude Measurement, http://webspace.utexas.eduedu/brizey9/www/4.html, page 2 of 3)
The problems that Thurston encountered such as ambiguity are covered here in Likerts statements, it is like Likert learned from Thurstones problems cab changed them in this measure of attitudes scale.
Likert then using the 1 to 5 method assigned values to each of the 1 to 5 outcomes, this was averaged.
Seiter and Hough (1970) said;
" Likert's method proved to be just as reliable and much simpler than Thurstones". (24/11/2003, Attitude Measurement, http://webspace.utexas.eduedu/brizey9/www/4.html, page2of 3)
Both Likerts and Thurstones methods are self reporting using a cognitive component.
The other measures of attitude include multiple indicator and indirect testing but there is also other conceptualisations of attitudes such as social representations, physiological measures of attitudes and semantic differential
(Osgood et al, 1957, Chapter 4)
Indirect tests " Utilize and illustrate psychological laws to a greater degree that direct attitude test and are thus more characteristic of measurement In the successful sciences where in yesterday crucial experiments are today routine measurement procedures".
(Campbell and Kidder , 1957, page 333.)
The focus of indirect testing is that that participants do not actually know what is being tested, bias are less likely in this situation as the participant does not know what the researcher is testing so the participant will not say what he/she thinks the researcher want to hear, it is less likely the participant will conform.
Indirect tests are deceptive and this could incorporate many ethical issues.
Kidder and Campbell (1957, page 336) quote;
" Characteristic of the indirect test is a façade. By this is meant a false assignment to the respondent which distracts him form recognising the tests true purpose and which provides him with a plan able reason for co-operating."
The main problems of this type of attitude measurement is it rises ethical issues as the research is being deceptive and misleading.
The problem with attitude measurements is peoples attitudes are exceptionally complex. The multiple indicator method takes this view , this also suggests;
"The communication of attitudes often involve a set of behaviours and dispositions of the subject" (24/11/2003, Attitude Measurement, http://webspace.utexas.eduedu/brizey9/www/4.html, page2of 3)
The multiple indicator test has two main variables that it takes into account characteristics of the person, his views, ethics, expressive style etc also his expectations how he acts etc
(Cook and Selitiz, 1964, page 220)
This method is more realistic as it takes into account individual differences, opinions, views etc.
Another problem with measuring attitudes is you do not know how the researcher perceive things, their understanding of words and sentences, a sentence could be ambiguous or individuals perceive the sentence different to the person nest to them, so then the problem is greater as the participant could mean and interpret sentences different.
(Osgood et al, 1957)
researched this, they focused on the meaning people give to a word or concept, thy said words have two meanings;
. Semantic or dictionary meaning or
. Connotative meaning.
(Hoff and Vaughan, 1995, Chapter 5)
From the evidence denoted it is obvious that measuring attitudes is always going to be a difficult and unpredictable test.
Our social representations and our social constructions, the ideas and opinions which we hold are moulded by what other people believe or say, this alone creates problems of biases in peoples attitudes.
People tend to conform in social situations, you answer by what you think other people would say or think, your inward attitude is not always tour outward attitude.
An important factor in all these methods mentioned is, are the tests reliable and valid.
Validity is a significant issue, who interprets is what is being measured an attitude?
"Critic would say that what is actually being measured is demand characteristics" (Hoff and Vaughan, 1995, Chapter 5)
Demand characteristics is when the participant respond to cues within the experimental situation for example experimenter behaviour
(Birch, A & Malim, T, 1998:825)
The actual definition of an 'attitude' is in itself questionable.
The self reporting method has problems such as participants know they are being tested so more likely to guess the response rather than honestly respond to their attitude, this then becomes unreliable . If someone has a undesirable attitude about something they may lie to conceal this, this consequently sabotages the testing.
As noted before indirect testing has ethical problems, very exploitive and deceptive.
Statistic is not going to calibrate peoples attitudes and beliefs there are feeling that are un-measurable, empathy etc, when and how do you distinguish all these things.
Further studies are definitely needed to improve the various methods that are about today. Peoples attitudes control our everyday life and it would be invaluable to accurately measure it.
References
Birch, A & Malim, T, (1998) "Introductory psychology, Hampshire, Palgrave.
Cardwell et al, (1996) "Psychology for A Level", London: Harper Collins Publishers
Retrieved 24/11/2003, Attitude Measurement, http://webspace.utexas.edu/brixey9/www/frame.html.
Hoff & Vaughan, (1995) Attitude Measurement, Chapter 4.