Consideraciones y observaciones, requisitos in dispensables y estilo

A series of questions in a variety of formats was used to provide the data. These included dichotomous items, multiple response, rating and Likert-scales, scenarios with

75 Section A of the questionnaire measured five sub-topics; demographics, religiosity,

experience with OTCs, label reading and perceived health condition. Label reading, although treated as an outcome of the socialisation process, was placed in this section as it related to experience with OTCs. As such it improves the continuity and flow of the questionnaire. Questions 1 to 10 of Section A consisted of a standard list of demographic and socio-economic items.

Measuring adolescents’ socioeconomic status was not a straightforward task. Research into understanding adolescents’ socioeconomic status has used various measures. While most studies used parents’ education and/or occupation (Marjoribanks, 1995; Torre et al., 2006; Wang & Kao, 2007; Yi, Kung, Chen, & Chu, 2008) as a proxy for adolescents’ socioeconomic status, some studies have also used pocket money (Currie, Elton, Todd, & Platt, 1997; Xie, Chou, Spruijt-Metz, Reynolds, & et al., 2007). This study used three variables to measure adolescents’ socioeconomic status; pocket money spent in a week, and parents’ occupations and education. To help respondents answer the demographic questions, most were presented in a categorical format. The amount of money spent per week, father’s job and mother’s job were in an open-ended format to obtain a more accurate response than a check list could offer. The answers were then collapsed into categories.

Religiosity was measured using a 6-point numerical scale for statements which were adapted from Worthington et al. (2003) and Mokhlis (2006a). Worthington et al. (2003) developed the Religious Commitment Inventory-10 in counselling psychology. The scale was tested in the United States and has strong estimated internal consistency, three-week

76 test-retest reliability, construct validity, discriminant validity and, criterion-related validity (Worthington et al., 2003). Mokhlis (2006a) adapted the Worthington et al. (2003) scale in Malaysia in his study on the effect of religiosity on shopping orientation and obtained good reliability (Cronbach’s alpha) as well. These two studies illustrate that the scale has been validated in places with different distributions of religious affiliations and in different research topic areas. As used in the present study, four of the five questions have a scale ranging from 0 (not at all true of me) to 5 (totally true of me), while one question utilised a scale from 0 (never) to 5 (every day).

The next six questions measured experience with OTCs. These included usage, purchase, and label reading. Respondents were given a definition of OTCs and some examples. A screening question asking whether respondents had used OTCs was presented.

Subsequently, reasons for OTCs usage were measured by a multiple-response checklist, adapted from previous OTCs’ research (Chambers et al., 1997; Covington, 2006; Holstein et al., 2004). Questions about the purchase of OTCs were in a multiple-response format as well. A “yes” or “no” response established whether the respondents read labels followed by a question to determine how carefully they read them. For this question, a scale ranging from 0 (brief scan) to 5 (every word) was used. The final part of this section asked about respondents’ health condition. Respondents were asked to indicate their level of health, ranging from 0 (poor) to 10 (excellent). To provide a finer indication of general health, respondents were then asked how many times they had been ill in the past 12 months.

Section B contained questions about consumer socialisation, beginning with interaction with family. Eight statements were included, of which two statements were general

77 statements about family interaction. The other six statements specifically focused on family interaction with regard to OTCs. This scale was developed in such way as to explore the degree to which the type of communication between the respondent and his or her parents was more socio or concept-oriented. All items were developed specifically for this study, but reflected the literature on family communication patterns (Chan & McNeal, 2003; Lachance et al., 2000; Mangleburg & Bristol, 1998; Moschis, Moore, & Smith, 1984) and OTCs. The original scale for family communication patterns was a 5-point Likert scale but was changed to a “yes” or “no” dichotomy for this study. The use of a short, precise, dichotomous scale helped students to make a quick and certain decision, considering the limited time given and to balance other more difficult questions at the end of the questionnaire. The decision to use the dichotomous format was made following the feedback from a pilot testing exercise.

The section continued with items measuring interaction with peers. Seven statements specifically focused on peers’ interactions regarding OTCs, whilst one general statement was also included. As with family interaction, all items measuring peer interactions regarding OTCs were developed specifically for this study, but reflected the literature on normative and informational peer influences (Bearden et al., 1989; Mangleburg & Bristol, 1998) and OTCs. Similar to interaction with family, the original scale for peer interaction was changed from a 5-point Likert scale to “yes” or “no” answers, for the reason noted above.

Section B continued with interaction with the media. Respondents were asked to indicate the media they would use if they wanted to find information about OTCs. Ten options were listed and respondents could report more than one source in a multiple-response

78 format. The media listed were selected on the basis of their relevance to consumer

socialisation (Ferle et al., 2000; Mehta & Keng, 1985; Moschis & Churchill, 1978) and from the OTCs’ literature (Hughes et al., 2002; John & Evans, 2000). This was followed by three questions about internet usage. Additional questions about internet use were included as the internet was a significant source of market information among adolescents (Singh et al., 2003).

Communication by the school about drugs, both illegal and OTCs, was measured using four statements specifically developed for this study. These used a “yes” or “no” dichotomy. This was followed by two questions measuring communication with others who may be consulted by adolescents in the consumption of OTCs. Both items were in a multiple-response format. The first question listed the other people who may be consulted by adolescents. The list was based on previous OTCs studies (Chambers et al., 1997; Neafsey, Jarrín, Luciano, & Coffman, 2007; Paddison & Olsen, 2008; Yousef et al., 2008). The second question listed the type of information given by these people, adapted from Portner (1991).

Section C comprised questions about knowledge and attitudes. To measure adolescents’ knowledge about OTCs, questions about marketplace knowledge and factual knowledge were presented. Marketplace knowledge refers to teens’ understanding of consumer- related factors such as stores, shopping and prices (Mangleburg & Bristol, 1998). Based on this definition, three statements were developed. Measurement of factual knowledge consisted of four questions based on those used by Moore et al. (2002). These items measured adolescents’ knowledge about OTCs and the potential consequences of use.

79 Questions about attitudes were divided into those about medicines in general (both

prescribed and OTCs) and about OTCs only. A 5-point scale was used for both questions. Five items were used to measure attitudes toward OTCs and seven items were used to measure attitudes about medicine in general. Most of these items were a reflection of the literature on OTCs (Covington, 2006; Hughes et al., 1999; Huott & Storrow, 1997; Wazaify et al., 2005).

Section D contained questions to measure behaviour towards OTCs; proper and improper use. For sensitive behaviours such as these, Malhotra, Hall, Shaw and Oppenheim (2002) suggested a third-person technique, phrased as if they referred to other people. Lee (1993) recommended using “short descriptions of a person or a social situation which contained precise references to what were thought to be the most important factors in the decision making or judgement making processes of respondents” (Alexander and Baker, cited in Lee, 1993, page 79). Based on this, four scenarios were used to measure impressions of proper and improper use. A 5-point scale was used for this measurement.

The final question was open-ended, designed to collect thoughts, observations or opinions of the respondents about adolescents and OTCs. The researcher was hoping for richer insight into the way young people perceived and acted with regard to OTC medication (Malhotra et al., 2002).