**6. Methodology**

### **6.1. Aim**

Wai et al. [40] found, in common with other studies including Gelber [41] and Elmaleh [42], that contact and knowledge factors influence the development of positive attitudes towards

Experts in attitude research generally divide methods to measure attitudes into direct and indirect approaches [50]. Direct methods involve the respondents being aware that they are participating in attitude measurement and typically involves self-report surveys. Due consideration should be given to the use of more subtle or indirect methods of assessing attitudes. People may often hold subtle forms of prejudice towards disability that may not be detected when using more direct methods that allow respondents to respond in ways they consider more socially appropriate [51] and therefore while it is now more socially appropriate for the public to espouse positive global attitudes towards disability than negative, specific attitudes, if investigated, may be found to be more negative [4]. This social desirability phenomenon, where it becomes more appropriate socially to express certain sentiments and attitudes, may not necessarily be reflected in behaviour. Ichheiser [4] was a social psychologist of the Austrian phenomenological tradition wrote in depth about attribution biases long before other theorists. He theorized that people have two aspects of personality, their visible outside that is exposed to pressure and control from others and an invisible inside. In other words, society misshapes the image of the other person by describing certain characteristics to their personality. He argued that attitudes and opinions are often based on the assumption people do have definite attitudes and opinions, however most people have confused ambiguous indefinite opinions and attitudes about many things,

Therefore, Ichheiser [4] maintained that the findings in attitude research seriously missed the point. The reason for this was that people expressed to use '*views in principle*' and '*views in fact*'. He suggested that views in principle are how people think they would act or how they think they should act when confronted by issues are events that are important. *Views in fact* determined a person's actions and reactions when exposed to certain events or issues. He posited that both views were genuine. This theory assists in explaining how members of society, for example, who live in areas where there were no community-homes express no reservations about having people with learning disabilities as neighbours [views in principle]; whereas views in fact surface when a community based home for people with learning disabilities is next door to their home. This latter attitude is referred to as not in my backyard (NIMBY)'effect. Ichheiser [4] theorized that to prevent distance confounding attitude results, immediate neighbours should be targeted in attitude research. Therefore, in researching attitudes, a micro-

Decisions on choice of instrumentation to measure attitude are always a compromise between the ideal and the practical. It is advisable, however, to use a previously developed validated instrument that has been used in similar attitude [27] research, as showcased in the exemplar, Methodology section. Other widely used instruments to examine attitudes towards people with learning disabilities include the *Attitudes towards Disabled Persons Scale* [ATDP, 52] and the *Scale of Attitudes towards Disabled Persons* [SADP, 53]. Both these tools assess attitudes from a

and their attitudes and sentiments are often still more confused [4].

neighbourhood has utility and applicability.

people with disabilities.

190 Learning Disabilities - An International Perspective

**5.1. Researching attitudes**

The aim of the current attitude research was to identify community attitudes to people with learning disabilities living in the community in Southern Ireland.

#### **6.2. Research design**

The current study utilized a comparative descriptive design, combining a random survey and a micro-neighbourhood sample. There are many research designs in attitude research and one of the authors would recommend to showcase an exemplar of attitude research is a comparative descriptive design when ascertaining attitudes to people with learning disabilities living in the community. This design has utility and practical application when endeavouring to compare the attitudes of two groups within the population under study. In line with the importance of researching attitude, the rationale behind using a micro-neighbourhood design and a simple random survey design was to ascertain true attitudes (views in fact as opposed to views in principle), as evidenced in the literature review.

This comparative descriptive design incorporates a micro-neighbourhood design [63; **Figure 1**] and a simple random survey design. A micro-neighbourhood is defined as surveying the houses surrounding a target house(s), that is, community-home for people with learning disabilities. The researcher using this design is required to survey two houses on either side of the target house on the same side of the street; the five closest houses across the street and the three closest houses behind the target house (*n* = 12). If the target house deviates from this definition in one respect or another, e.g. if houses did not exist behind the target house, the sampling frame should be reduced rather than extended.

#### **6.3. Research sample**

A 5% margin of error (the error in estimating a proportion with 95% confidence) is the one most commonly used in research, more especially if, a priori, no such research had been previously carried out. One method of calculating the sample size (n) required for a given some population (N) is to substitute the values in this equation [64, 65]. Five percent margin of error = 1.96 (√2500 / N) × (1−n / N); however, statistical calculation using statistical software produces a faster and more accurate calculation. The population in the chosen electoral areas was (29,490) therefore 400 was estimated as a representative sample and allowed for a 5% margin of error. A simple random selection was used to proportionally select the sample from each of the electoral areas in the city, using the electoral register

**Figure 1.** Comparative descriptive design.

(Area 1 = 8191 ; Area 2 = 10,453; Area 3 = 10,846) total (n = 200) (N = 29,490). The target microneighbourhood houses (n = 158) were excluded from the random selection. Therefore, the total sample consisted of 358 houses.

#### **6.4. Recruiting and retaining sample**

It has been noted that very few people refuse to take part in social research [65]. Therefore, there is a need for extra care when seeking access for research purposes to ensure fully informed consent obtained and full disclosure is made regarding the operationalization of the research process. A letter was sent to prospective respondents prior to the survey, explaining the rationale, aims and objectives of the research. The research information sheet explained how the final report would benefit services in the planning of future of services in the community for people with learning disabilities. The respondents were informed that the findings of the research would be made available to them if such a request was made.

#### **6.5. Ethical approval**

**6.2. Research design**

192 Learning Disabilities - An International Perspective

**6.3. Research sample**

**Figure 1.** Comparative descriptive design.

The current study utilized a comparative descriptive design, combining a random survey and a micro-neighbourhood sample. There are many research designs in attitude research and one of the authors would recommend to showcase an exemplar of attitude research is a comparative descriptive design when ascertaining attitudes to people with learning disabilities living in the community. This design has utility and practical application when endeavouring to compare the attitudes of two groups within the population under study. In line with the importance of researching attitude, the rationale behind using a micro-neighbourhood design and a simple random survey design was to ascertain true attitudes (views in fact as opposed

This comparative descriptive design incorporates a micro-neighbourhood design [63; **Figure 1**] and a simple random survey design. A micro-neighbourhood is defined as surveying the houses surrounding a target house(s), that is, community-home for people with learning disabilities. The researcher using this design is required to survey two houses on either side of the target house on the same side of the street; the five closest houses across the street and the three closest houses behind the target house (*n* = 12). If the target house deviates from this definition in one respect or another, e.g. if houses did not exist behind the target house, the sampling

A 5% margin of error (the error in estimating a proportion with 95% confidence) is the one most commonly used in research, more especially if, a priori, no such research had been previously carried out. One method of calculating the sample size (n) required for a given some population (N) is to substitute the values in this equation [64, 65]. Five percent margin of error = 1.96 (√2500 / N) × (1−n / N); however, statistical calculation using statistical software produces a faster and more accurate calculation. The population in the chosen electoral areas was (29,490) therefore 400 was estimated as a representative sample and allowed for a 5% margin of error. A simple random selection was used to proportionally select the sample from each of the electoral areas in the city, using the electoral register

to views in principle), as evidenced in the literature review.

frame should be reduced rather than extended.

Ethical approval was sought from the university where the research was supervised and from main service provider for people with learning disabilities in a city in the South East of Ireland.

#### *6.5.1. Anonymity and confidentiality*

Assurances of confidentiality were conveyed in writing both in the research information sheet that was left with respondents and in the letter that was sent to participants prior to the survey being carried out. Participants were reassured that information would be analysed as group data so that individuals could not be identified by their responses and questionnaires were assigned number codes for statistical purposes only.
