Anda di halaman 1dari 5

INTRODUCTION Questionnaires constitute the, most prevalent method of information collection a mong the communication methods used.

Both structured and unstructured questionna ire are in use in mar eting research. The unit discusses various issues regardin g questionnaire, their types as well as the steps in designing a questionnaire. A sample questionnaire is also included to familiarise you with the structure an d design of questionnaire. 6.2 QUESTIONNAIRE: FORMAT AND ADMINISTRATION A questionnaire is a standardised format of data. collection. It is normally use d when the data is collected from a large population about their awareness, atti tudes, opinions, past and present behaviour. While there is no standard procedur e to construct a questionnaire, the problem, situation and experience of the res earcher greatly influence the process. The respondents may be as ed questions wi thout any formal questionnaire or with a pre-planned properly framed questionnai re. There could be a number of variations in the use of questionnaire method of data collection. They could vary in terms of its `format' and also in terms of t he way it is `administered'. Questionnaire Format Questionnaire format depends upon the amount of structure and diguise required d uring data collection, a) Structure At the time of fronting the questionnaire the researcher must appropiiately dete rmine the degree of structure to be imposed on the questionnaire. A highly struc tured questionnaire is one in which the question to be as ed and the responses p ermitted are explicitly pre-specified. On the other hand in a non structured que stionnaire the questions to be as ed are ept flexible in their own words and al so the respondents are allowed to answer the questions in a manner they li e. Th e response pattern may vary from open-ended to closed-ended. In open-ended quest ion the respondent is free to choose the possible response, where as in the clos ed ended from, the researcher pre specifies certain options and the respondent i s allowed to choose the alternative(s) from the given options. For example, the structure of these two forms of response will be as follows: Open-ended : What brand of shampoo do you use? Close-ended : Mention the brand of shampoo you use from the list given below: ( ) Ponds ( ) clinic ( ) Tiara ( ) Palmolive b) Disguise Disguised questions is one where purpose is not made obvious to the respondents and is as ed in an indirect manner. Non-disguised questions, on the other hand, are ones which are direct and the purpose of as ing them is nown clearly is the respondents. Disguised questions are used in the conditions when the issues con cerned are such that respondents may not give correct answer to direct questions . Based on the above discussion, questionnaires could be classified into for categ ories. Structured, non-disguised questionnaire Structured, disguised questionnaire Non-structured, non-disguised questionnaire Non-structured, disguised questionnaire Structured, non-disguised questionnaires are very popular in mar eting research studies. These are more applicable when large sample sizes are there. Non-struct ured, non-disguised questionnaires, on the other hand, are used when a freehand is to be provided to the respondents so that an in-depth information on the subj ect could be solicited e.g. in industrial mar eting research wherein number of r espondents would also be low. Non-structured, disguised questionnaires are mainly used in `motivation research '. 'Wore Association Test', 'Sentence Completion Test', `Thematic Appreciation T est', 'Cartoon Test', etc. may be used in this category, Structured, disguised q uestionnaires are more appropriate where responses are required towards certain

sensitive issues li e attitude towards aids patients, abortion etc. Questionnaire Administration: So far we have discussed about different formats of questionnaires. However, the questionnaire method may also vary depending 'on the way it is administered: Th ese could be broadly classified into three different categories. i) `Personal interview', wherein there is a face to face interaction between int erviewer (s) and respondents (s). ii) 'Telephone survey`, in which survey is conducted over phone i.e. unli e pers onal interview there is only a voice contact. iii) `Mail survey', as the name suggests, is conducted through mail and as such there are no interviewers. 6.3 STEPS INVOLVED IN DESIGNINING A QUESTIONNAIRE Step 1, Preliminary decision a) Assess what information will be sought after a thorough scanning of secondary sources of data. b) Determine the target respondent. Step 2. Decide on the type of questionnaire and method of administration. Step 3. Evaluation of question content. Before including a question in the sched ule, examine whether. This question is really essential. The respondent can understand the question. (i.e. is it too technical, ambiguous , or advanced for the target respondent?) The respondent can answer the question. (Say, the respondents possess sufficient nowledge. As such, it is better not to as too much of factual data or about p ast history.) The respondent will answer the question. Specially, if it invades into one's pri vacy or it requires too much effort to answer, then they usually refuse to coope rate. Step 4. Chec question phrasing. For Example Do words have ambiguity in meaning? Are there any implied alternatives in the question? Are there some assumptions to be made to answer the question'? , Will the respondents approach the question from the same frame of reference as d esigned by the researcher? Step 5. Determine form of response to each question: The response format may be open or closed ended. In general, the type of respons e format will depend on the objective of the research, nature of data to be coll ected and analysis to be performed. Step 6. Determine sequence of questions. Use simple and interesting opening questions; Design branching questions with care; Put the questions in logical manner; As for classificatory data at the end. Step 7. Assess the physical layout of the questionnaire. Naturally, the question naire must be printed properly; put in elegant form and facilitate handling. Step 8. Pre-test the questionnaire. It is normal practice to pretest a questionnaire on a small number of target res pondents. The pretest is done to assess both individual questions and their sequ ence of response pattern. Accordingly, a researcher must revise questions which cause problems. While developing a questionnaire, researcher must Use simple words in framing the questions; Avoid ambiguous- questions; Avoid implicit alternatives; Avoid questions that require too much memory recall and calculation; Avoid double barreled questions; Ideally, a questionnaire should first secure' some basic information to get the respondent's cooperation and gradually try to collect more information about the phenomenon of interest.

It is easier to administer a multiple choice response categories that requires o ne simple tic . 7 Steps to Prepare Data for Analysis Tuesday, March 2, 2010 by Richard Pin We researchers spend a lot of time interviewing our clients eeds. Then we go about carefully creating a plan to collect e most useful. Having done that, the appropriate instrument that will generate data that can ultimately be transformed this up-front wor necessitates and lot of time and effort. But sooner or later we will have collected data and need to of data preparation.

So what is involved in data preparation? There are several simple, but sometimes overloo ed steps, required to properly prepare data. They are: Questionnaire chec ing: Questionnaire chec ing involves eliminating unacceptable questionnaires. These questionnaires may be incomplete, instructions not follow ed, little variance, missing pages, past cutoff date or respondent not qualified . Editing: Editing loo s to correct illegible, incomplete, inconsistent and ambigu ous answers. Coding: Coding typically assigns alpha or numeric codes to answers that do not a lready have them so that statistical techniques can be applied. Transcribing: Transcribing data involves transferring data so as to ma e it acce ssible to people or applications for further processing. Cleaning: Cleaning reviews data for consistencies. Inconsistencies may arise fro m faulty logic, out of range or extreme values. Statistical adjustments: Statistical adjustments applies to data that requires w eighting and scale transformations. Analysis strategy selection: Finally, selection of a data analysis strategy is b ased on earlier wor in designing the research project but is finalized after co nsideration of the characteristics of the data that has been gathered. Not all of these steps occur in every mar et research study. But as situations d ictate, none of these steps should be overloo ed in the name of expediency or ec onomy. Later articles will drill down in to the details of these important steps in data preparation. Chapter 2 Exploratory Factor Analysis 2.1 Objectives The primary objectives of an EFA are to determine 1. The number of common factors inuencing a set of measures. 2. The strength of the relationship between each factor and each observed measur e. Some common uses of EFA are to Identify the nature of the constructs underlying responses in a specic content ar ea. Determine what sets of items \hang together" in a questionnaire. Demonstrate the dimensionality of a measurement scale. Researchers often wish to develop scales that respond to a single characteristic. Determine what features are most important when classifying a group of items.

to determine their n the data that will b is carefully crafted into nowledge. All And well it should! start the grunt wor

Generate \factor scores" representing values of the underlying constructs for us e in other analyses. 2.2 Performing EFA There are seven basic steps to performing an EFA: 1. Collect measurements. You need to measure your variables on the same (or matc hed) experimental units. 2. Obtain the correlation matrix. You need to obtain the correlations (or covari ances) between each of your variables. 3. Select the number of factors for inclusion. Sometimes you have a specic hypoth esis that will determine the number factors you will include, while other times you simply want your nal model to account for as much of the covariance in your data with as few factors as pos sible. If you have measures, then you can at most extract factors. There are a number of methods to determine the \optimal" number of factors by examining your data. The Kaiser criterion states that you should use a number of factors equal to the number of the eigenvalues of the correlation ma trix that are greater than one. The \Scree test" states that you should plot the eigenvalues of the co rrelation matrix in descending order, and then use a number of factors equal to the number of eigenv alues that occur prior to the last major drop in eigenvalue magnitude. 4. Extract your initial set of factors. You must submit your correlations or cov ariances into a computer program to extract your factors. This step is too complex to reasonably be done by hand. There are a number of dierent extraction methods, including maximum li elihood, princip al component, and principal axis extraction. The best method is generally maximum li elihood e xtraction, unless you seriously lac multivariate normality in your measures. 5. Rotate your factors to a nal solution. For any given set of correlations and n umber of factors there are actually an innite number of ways that you can dene your factors and sti ll account for the same amount of covariance in your measures. Some of these denitions, however, are easier to interpret theoretically than others. By rotating your factors you attempt to nd a factor so lution that is equal to that obtained in the initial extraction but which has the simplest interpreta tion. There are many dierent types of rotation, but they all try ma e your factors each highly responsive to a small subset of your items (as opposed to being moderately responsive to a broad set). There are two major categories of rotations, orthogonal rotations, which produce uncor related factors, and oblique rotations, which produce correlated factors. The best orthogonal rotatio n is widely believed to be Varimax. Oblique rotations are less distinguishable, with the three most c ommonly used being

Direct Quartimin, Promax, and Harris-Kaiser Orthoblique. 6. Interpret your factor structure. Each of your measures will be linearly relat ed to each of your factors. The strength of this relationship is contained in the respective factor loading, produced by your rotation. This loading can be interpreted as a standardized regression coeci ent, regressing the factor on the measures. You dene a factor by considering the possible theoretical constructs that could b e responsible for the observed pattern of positive and negative loadings. To ease interpretation y ou have the option of multiplying all of the loadings for a given factor by -1. This essentially rever ses the scale of the factor, allowing you, for example, to turn an \unfriendliness" factor into a \friendline ss" factor. 7. Construct factor scores for further analysis. If you wish to perform addition al analyses using the factors as variables you will need to construct factor scores. The score for a given factor is a linear combination of all of the measures, weighted by the corresponding factor loading. Sometimes factor scores are idealized, assigning a value of 1 to strongly positive loading s, a value of -1 to strongly negative loadings, and a value of 0 to intermediate loadings. These factor score s can then be used in analyses just li e any other variable, although you should remember that they wi ll be strongly collinear with the measures used to generate them.

Anda mungkin juga menyukai