The Concept of Questionnaire Design by Highlighting the Major Steps in Avoiding Mistakes - Essay Example

Introduction A researcher has to develop a good questionnaire depending on research methodology in order to come out with a good research. The purpose of a questionnaire is determined by the information the investigator needs. This information could be either qualitative or quantitative in cases where the researcher wants to quantify hypotheses and the information has to be analyzed statistically (Bogdan and Biklen, 2008). A well-developed questionnaire should be made according to the research objectives stated. This might seem understandable, but several researches omit significant aspects correct to inadequate preliminary work, and do not sufficiently probe specific issues owing to poor identification. To a clear degree, several of these are unavoidable. Every research is compelled to leave certain questions unrequited and provide a necessity for additional research, but the aim of good questionnaire plan is to control these difficulties (Campbell, 1996). The questionnaire should achieve complete and the most accurate information as possible. The questionnaire designer needs to authorize that participants entirely comprehend the inquiries, and are not prone to reject to answer, recline the interrogator, or try to hide their approaches. Campbell (1996) indicates that a suitable questionnaire is prearranged and expressed to encourage participants to offer correct, unprejudiced, and complete report. A well- developed questionnaire should be easy for participants to give the essential report and for the researcher to document the answer and it should be organized so that real analysis and explanation are conceivable (Davies and Dodd, 2002). A good questionnaire would have the researcher brief and exact and be so agreed that the participants(s) remain involved throughout the research. In research methodology, there are nine steps considered while developing questionnaires. These steps are presented below. Determine the information needed It should be made clear that one does not begin by composing questions. The primary step is to determine the entities that one requires to know from the participants in order to understand the study's objectives. One may previously have an idea concerning the kind of data to be gathered, but supplementary help can be attained from secondary information- previous quick rural assessments, and exploratory investigation (Davies and Dodd, 2002). Concerning secondary information, the researcher must be conscious of what effort has been put on the same or related complications in the past, what features have not yet been assessed, and how the current study questionnaire can develop on what has previously been determined. Moreover, brief introduction with unofficial interviews with target research participants will give a hint of authenticity that may assist clarify philosophies about what data is required (Davies and Dodd, 2002). Outline the target study participants At the beginning, the investigator must describe the populace around which they need to simplify from the sample information to be gathered. For instance, in marketing inquiry, researchers over and over again have to choose whether they must include only current users of the general produce category or whether to incorporate non- consumers. Secondly, investigators have to extract up a selection framework. Thirdly, in planning the questionnaire, researchers must put into account influences such as the age, level of education, among others of the target research participants. Select the method(s) of accomplishment your target participants The approach of contact will affect not only the items the researcher is capable to inquire, but also the wording of those item questions in the questionnaires. The main techniques available in review research are individual interviews, grouping or focus interviews, emailed questionnaires and phone interviews. Select question content Researchers must continuously be ready to ask, "What is the importance of this question in my study?" The attraction to include issues without analytically evaluating their role towards the success of the research purposes, as they are stated in the study proposal, is remarkably strong (Charles, 2005). No item or question ought to be incorporated unless the information it gives increase to is openly of use in examining one or more of the hypotheses created in the study design. There are only two cases when seemingly “superfluous " questions may be incorporated- inaugural questions that are comfortable to answer and which are not observed as being " intimidating ", and/or are viewed as being motivating, can importantly assist in attainment the respondent's participation in the study and aid to ascertain a rapport. Expand the item question wording Research questions can be categorized into three methods, for instance, closed, open-ended, and open response-option queries. Closed questions have a number of significant advantages, which include the following; It offers the respondent with a comfortable method of point to his response – they do not have to contemplate about how to express his response. It makes the study participants to depend less on remembrance in responding a question. Answers can be easily categorized, making examination much uncomplicated. It allows the respondent to state the answer groupings most appropriate for their reasons. Place questions into a significant order and setup Introductory questions must be straightforward to answer and not intimidating to the research respondents (Charles, 2005). The initial question is critical because it is the participants’ first contact to the interview and puts the mood for the environment of the task to be achieved. If they discovered the initial question complex to understand, or was past their knowledge and understanding, or humiliating, they are prone to break off instantly. If, in contrast, they discovered the initial question comfortable and enjoyable to answer, they are stimulated to carry on. Questions must stream in some sense of psychological direction, so that one directs the next (Dillon, Madden and Firtle, 2004). Questions on one theme, or one specific aspect of a theme, should be categorized together. Research respondents might find it confusing to keep ever changing from one theme to another, or to be questioned to revisit to some theme they believed they gave their views about previously. Research respondents become uninterested quickly and fidgety when asked analogous questions for an hour or more (Neuman, 2001). It frequently improves reaction, therefore, to change the respondent's undertaking now and then. An open-ended item provided frequently offers much-needed assistance from a long sequence of questions in which research participants have been impounded to control their replies to already coded groupings. Questions including showing tags / images to participants can assist develop the pace and enlarge attention (Campbell, 1996). Concluding questions It is normal for a research respondent to turn out to be increasingly uninterested to the inquiry form as it approaches the end. This is due to displeasure or fatigue; they may offer thoughtless responses to the concluding items. That item is important and should be included in the initial questions in the questionnaire. Theoretically, delicate questions should be put to the end, to prevent respondents from getting demotivated before significant information is gathered (Neuman, 2001). In forming the questionnaire, the investigator or researcher should give particular consideration to the appearance and layout of the question form the situation. The interviewer’s role needs to be put as clear as possible. Items in the questionnaire be obviously worded and response choices clearly recognized (Campbell, 1996). Prescribed descriptions and clarifications should be offered (Campbell, 1996). This confirms that all correspondents control the inquiries repeatedly. Throughout the interview procedure, the researcher can reply /clarify respondents’ questions (Campbell, 1996). Sufficient writing gap should be permitted to put open-ended responses, and to consider for variances in writing between the respondents (Campbell, 1996). Physical outlook of the questionnaire The physical outlook of a questionnaire has an important impression upon both the amount and value of marketing information attained (Bogdan and Biklen, 2008). The amount of data is determined by a summary of the reaction rate. Ill- planned questionnaires can indicate an impact of complications such as time wastage. The physical outlook of the questionnaire with unreasonably confusing designs makes it more problematic for respondents’, or research participants in the case of self- achievement questionnaires, to conclude this task correctly, thus, influence information worth. Consideration to just an insufficient basic particulars can have an unreasonably advantageous influence on the information obtained in a questionnaire (Bogdan and Biklen, 2008). Pre-test the research questionnaire It is vital to pre-test the questionnaire before it is used in obtain information from the study participants to identify any errors that need amending (Bogdan and Biklen, 2008). The aim of piloting the questionnaire is to control: · Whether the items as they are worded will produce the anticipated results · Whether the items in the questionnaire have been put in the best direction · Whether the items are comprehended by all categories of respondent · Whether supplementary or specifying items are desirable or whether some items should be removed · Whether the directions to interviewers are sufficient Once a questionnaire has been exposed to a systematic pilot test, the ultimate form of the items and questionnaire will have changed into its last form. The last part will be gathering and sequencing the items into a suitable order, totaling questions, and introducing interviewer directions (Bogdan and Biklen, 2008). Bogdan and Biklen (2008) say that research questions and objectives needs to address these factors: (a) the sorts of quantitative research queries a researcher is trying to find a solution; (b) the variables the researcher wants to measure, influence, and/or manipulation; (c) the researcher should look at structure of research questions; and (d) he or he should use research questions as divergent to research hypotheses. Scaling can be well defined as the course of calculating the quantitative characteristics of individual or non-concrete concepts (Campbell, 1996). Most writings on marketing research elucidate the four concentrations of measurement: nominal, ordinal, interval, and ratio (Campbell, 1996). Nominal scales This, the simplest of measurement scales, categorizes individuals, businesses, products, products or other units into groups where no order is suggested (Campbell, 1996). Actually, it is often denoted to as a tentative scale. It is a structure of arrangement and does not put the unit along a range (Campbell, 1996). It contains a simply tally of the rate of the cases allocated to the various classes. Ordinal scales These scales involve the placing of individuals, opinions, or items along the range of the typical being grouped (Campbell, 1996). Interval scales The cardinal or interval scale has equal components of measurement, hence making it probable to translate not only the peace of scale scores but also the space between them (Clont, 2002). However, it must be documented that the zero point on an interval scale is random and is not a true zero. This has consequences for the type of data exploitation and analysis one can carry out on information collected (Clont, 2002). Ratio scales The maximum level of measurement is a ratio scale (Cohen, Manion and Morrison, 2004). This has the characteristics of an interval scale unruffled with a fixed derivation or zero point. Instances of variables that can be measured using ratio scaled contain weights, times and lengths. Ratio scales allow the researcher to evaluate both dissimilar scores and the comparative magnitude of counts (Cohen, Manion and Morrison, 2004). For example the dissimilarity between 5 and 10 seconds is the same as that between 10 and 15 seconds, and 10 minutes is double as long as five minutes. A measuring device links more one article in order to quantity a perception that is too complicated to be computed simply, or where one entry is a composite measure (Clont, 2002). Composite measures include both Scales and index measures (Clont, 2002). Quite a lot of concepts, like self-esteem, information, fulfillment with service industries, or need for facilities use composite measures. Normally dissimilar items are combined to generate a single count that is used to denote the value for that notion (Clont, 2002). According to Creswell and Miller (2000), an index is a composite degree that condenses and grades several exact observations, as well as characterizes some overall measurement. An index is an accretion of totals from a variability of individual entries. Three criteria for good measurement include the following; 1. Reliability 2. Validity (Face or content validity, Criterion validity, Construct validity, Convergent validity, Discriminant, validity) Reliability is the level to which a measurement tool produces unchanging and reliable results, whereas, validity denotes how well an assessment quantities what it is supposed to measure (Davies and Dodd, 2002). An attitude is a psychological state concerning feelings, beliefs values, and personalities to put-on in a clear way (Crocker and Algina, 2006). There are three elements of attitude - intellectual component, affective constituent, and behavioral constituent (Crocker and Algina, 2006). Attitudes can be measured using the following approaches; Bogardus' Social-distance Scale, Likert's Scale, Thurstone's Scale, Guttman's Scalogram and the Semantic Differential (Dillon, Madden and Firtle, 2004). Looking at the Likert scale, several opinion reports are collected, corrected and then specified to a group of themes to rate the reports on a five-point scale: for example 1 strongly agree; 2 agree; 3 undecided; 4 disagree; and 5 strongly disagree (Dillon, Madden and Firtle, 2004). An example of a latent construct is presented below. Respondents Frequency Adequate About adequate Inadequate Not sure Business Students 10 12 14 0 Academic Staff 2 1 5 4 Senior Administrators 2 3 9 1 Total 14 16 28 5 According Healy and Perry (2000), ranking undertakings are best achieved over a small quantity of objects. Participants requested to rank twenty-five different brands of automobiles from their greatest preferred to their slightest preferred most likely could designate with certain degree of consistency their most preferential, their second most desired, and their least favored. Whichever rankings- between the first two and lowest one- are prone to be unreliable. To make the ones in between positions dependable, researchers characteristically limit participants to ranking only a rare thing; as a rule of flick through (Eiselen, Uys and Potgieter, 2005). Sorting is similar to ranking on the whole. The benefit to sorting as a powered task is that researchers can request people to sort countless more objects than researchers can request them to rank (Eiselen, Uys and Potgieter, 2005). Therefore, sorts tend to be more consistent than ranking for a big number of substances. Measurement can be achieved using homogeneous scales or complete custom calculated scales. Harmonized scales may be chosen in case of calculating concrete substances (Joppe, 2000). Developing modified scale is necessary in the situation where researcher needs to quantify more non-concrete and multifaceted hypothesis like the consumer attitudes concerning a new product presented in the marketplace. In this circumstance, consistent scales may not be existent. This state permits the expansion of modified scales (Joppe, 2000). Joppe (2000) says that selection or creation of a measurement scale demands decision in the following six important areas; 1. Research objective: The researcher may have general objectives to measure the characteristics of study participants. 2. The three types of measuring scales must be put into consideration. This could be in the form of ranking order, rating, and categorization. 3. The researcher must consider the degree of preference where preference involves preference evaluation and preference measurement. 4. The properties of data must be viewed regarding measurement scales where data can be classified as nominal, ordinal, interval, and ratio. Data analysis depends on measurement scale. 5. The measurement scales can be classified as either multidimensional or one-dimensional. 6. When constructing scale these approaches are considered, consensus, arbitrary, item analysis, factoring and cumulative. Conclusion In business research, a researcher has to primarily define what they want to measure, how this measurement will be done and also the aspect that requires measurement (Patton, 2002). Measurement scales are usually categorized into four, which are, nominal scale, ordinal scale, interval scale, and ratio scale. The major criteria for testing the appropriate measurement include reliability, sensitivity, and validity, relevance, and generalizability. As evidenced in the above research, it is valid to argue that preparing a good questionnaire design involves a myriad of steps so as to attain precision. It is the role of the researchers to ensure that they are conversant with all the steps towards completion of a good research design. The paper has attempted to evaluate the concept of questionnaire design by highlighting the major steps in avoiding mistakes in a questionnaire design. Next in line was an assessment of the measure concepts of a research design. Lastly, attitude measurements were scrutinized in an endeavor to understand the perceptions that business researchers hold regarding attitudes. References Bogdan, R. C. & Biklen, S. K., 2008. Qualitative research in education: An Introduction To Theory And Methods 3rd ed.. Needham Heights, MA: Allyn & Bacon. Campbell, T., 1996. Technology, multimedia, and qualitative research in education. Journal of Research on Computing in Education, 309, 122-133. Charles, C. M.,2005. Introduction to educational research 2nd ed.. San Diego, Longman. Clont, J. G., 2002. The concept of reliability as it pertains to data from qualitative studies. Paper Presented at the annual meeting of the South West Educational Research Association. Houston, TX. Cohen, L., Manion, L., Morrison, K., 2004. Research methods in Education.5th Edition. Routledge-Falmer: London Creswell, J. W. & Miller, D. L., 2000. Determining validity in qualitative inquiry. Theory into Practice, 393, 124-131. Crocker, L., & Algina, J., 2006. Introduction to classical and modern test theory. Toronto: Holt, Rine Hart, and Winston, Inc. Davies, D., & Dodd, J., 2002. Qualitative research and the question of rigor. Qualitative Health Research, 122, 279-289. Dillon, W. R., Madden, T. S and Firtle, N. H., 2004. Marketing Research in a Marketing Environment, 3rd edition, NY: Irwin. Eiselen, R., Uys, T., Potgieter, N. 2005. Analysing survey data using SPSS13: A Workbook. University of Johannesburg. Healy, M., & Perry, C. 2000. Comprehensive criteria to judge validity and reliability of qualitative research within the realism paradigm. Qualitative Market Research, 33, 118. Joppe, M., 2000.The Research Process. Retrieved from http://www.ryerson.ca/~mjoppe/rp.htm. Access date: 31st January 2013. Neuman, W.L., 2001. Social Research Methods: Qualitative and Quantitative approaches, 4th edition. Massachusetts: Allyn and Bacon. Patton, M. Q., 2002. Qualitative Evaluation and Research methods 3rd ed. Thousand Oaks, CA: Sage Publications, Inc. Read More