Research projects which use a questionnaire to collect standardized data from a large number of people. Can be either Population or Sample surveys. Sample surveys are the most common The collection of standardized data requires that the same questions be given to all respondents in the same order.
Factual Surveys ² Use to collect descriptive information. Example, Population census, The Survey of Living Conditions and The Labour Force Survey. Attitude Surveys ² Carried out by opinion poll organizations, market researchers, etc. Explanatory Surveys - Used to test hypotheses or to test and develop theories.Common to all types, is the use of the Questionnaire as the instrument of data collection
A questionnaire is a collection of questions and /or statements that is designed to collect information on a particular topic. It is an instrument used by researchers to convert into data, information directly given by respondents. In essence, it provides access to what is inside the persons head
The questionnaire facilitates themeasurement of what a person: knows - knowledge, information likes & dislikes - values, preference thinks - attitudes, beliefs experiences - past & presentIt is a useful alternative when direct observation is not possible.
This approach to data collection requiresthat the respondent: co-operates in the completion of questionnaire tells what is, instead of what he thinks ought to be, or what he imagines the researcher would like to hear. knows how he feels or thinks in order to report.It is possible therefore for the questionnaire to measure notnecessarily what a person likes, believes or thinks but whathe/she indicates in these regards.
The researcher must, therefore, pay attentionto the following factors: the respondent will have a tendency to show self in good light. he/she may be unduly helpful by providing answers he thinks the researcher wants instead of telling it like it is. he/she may not be able to provide answers to the questions posed - out of ignorance etc.
Direct versus indirect (Specific vs. Non ²Specific) a. Do you like your job? - direct (specific) b. How do you feel about your job? - indirect (non-specific) a. How you feel about teacher A? - direct (specific) b.How do feel about class taught by teacher A? - indirect (non-specific)Direct or specific questions may cause respondent tobecome guarded or cautious and give less than honestanswers. Non-specific ones lead to desired informationwith less alarm.
Fact versus opinion a. What kind of car do you drive? b. Do you prefer Japanese or American?Factual questions do not always solicit factual answersbecause: i. faulty memory ii. conscious desire to create a certain impressionNor do opinion ones always solicit honest opinions.Respondents are normally inclined to provide sociallydesirable answers.
Questions versus Statements - Can be a direct question as those types above (requiring a direct answer) or a statement requiring an optional response. Predetermined versus Response Keyed Questions - Answer all vs. answer those that are relevant.
5. Do you drink alcoholic beverages? 1. Never 2. Occasionally 3. Frequently 4. Always(If never, go to 6 and then terminate. Otherwise, skip to 7 and continue)6. Why don·t you drink alcoholic beverages? 1. Religious reasons 2. Health reasons 3. Others (Specify) ______7. When you drink, which of the following are you most likely to have? 1.Rum 2. Beer 3. Stout 4. Wine 5. Others (Specify)________
Provide respondent with possible answers and ask him/her to choose the most appropriate option. When the closed-ended format is used, the researcher should be guided by the following: - Response categories provided should be exhaustive - Response options should be mutually exclusive - There should be clear instruction to select the best answerThis format is ¶respondent friendly· and facilitates greater ease in theprocessing of data, since it can be transferred directly to computer. Ithowever, limits the possible answers to those thought of by theresearcher.
Researchers ask questions and allow respondents to provide answers Exert control only in regard to the questions asked and the time and space provided. Respondents give own answer, rather than just agreeing with those given. Format offers the respondent more flexibly
Responses must be coded before processing - The coding process can be time consuming and can be quiet technical. It requires the researcher to accurately interpret the meaning of respondents give to responses. There is always the possibly of misunderstanding and researcher·s bias. Respondents quite often provide answers that are irrelevant to researchers intent.
This is transitional mode between structured and unstructured mode. Respondents generate, rather than choose answers Responses are, however, limited in range and length - often a single word or short phrase Example: What is your fathers occupation? The very wording of the question restricts the number of possible responses and the number of words.
Tabular Responses - Fill response into a table. A very convenient way of organizing complex responses. Scaled Response - A structured response form. Respondents are asked to express endorsement or rejection of a given statement. Example: The Likert Scale Ranking response ² Respondents are given some statements, etc. and asked to rank according to some criteria. Checklist Response - Respondents choose all possible answers from a number of options given to him
In constructing the Questionnaire, the researchershould always consider the following factors:Format ² Wording Precision ² Questions should be clear and unambiguous Concision ² Items should be as short as possible Relevance ² Question should all be relevant and necessary Double-barreled Questions ² Each question should should attempt to measure only one variable at a time Biased Items/Terms ² Should not use leading questions Negative Items ² Questions should be in positive form Abbreviations and Jargons ² These should always be avoided
Format ² Layout Uncluttered ² Items should be well-spaced/ spread- out Order ² Items should flow in a logical order. The ordering of questions affects the quality of responses Length ² Should not be too many items ² Instrument shouldn·t be too long Personal Information ² Request only when required Instructions ² Always provide adequate instructions ² both general and specific.
Three main Categories, based on the approach use in the completion of the questionnaires: - Mail Questionnaire - Face-to-Face Interviews - Telephone Interviews
Postal services are utilized in the distribution and return of instrument Classical approach is to send questionnaire accompanied by a letter of explanation and self- addressed stamped envelope. Respondents asked to complete and return within a specified time.
Cost ² the cheapest Time ² the slowest Degree of obtrusion ² the least obtrusive Specificity ² the least definite/certain Literacy ² absolutely necessary Response rate ² the lowest
This problem of a low response rate can have implications forgeneralization. Hence the need to incorporate measures in thedesign to ensure the highest possible response rate. A commonapproach is the use of follow-up mailing. Can take two forms: Reminder only to non- respondents Letter to all, thanking those who have responded and a reminder to those who haven·t as yet. Always include include a copy of the questionnaireEven with all practical measures, a 100 % response rate willnever be achieved. Researchers must decide, in advance, whatrate is considered acceptable ² that is, the minimum rate thatwill not introduce response bias.
Most popular form in the Caribbean. Interviewers askquestions and record answers as given. Most obtrusive form,so special attention must be paid to interviewers· competence,behaviour and appearance.Interviewers should,therefore, always display a pleasant and professional demeanor be familiar with questionnaire and research area. follow wording and format of questionnaire exactly record responses exactly as given use probing questions cautiously. be properly trained.
Cost ² the most expensive Time ² the most time consuming Degree of obtrusion ² the most obtrusive Specificity ² the most specific Literacy ² not necessary Response rate ² the highest
Most convenient approach, but with obviousBiases. Will have access only to those with telephones who are listed in the directory are available at the time of the interview
Cost ² Time ² Degree of obtrusion ² Specificity ² Literacy ² Response rate ²
1. Identification and specification of variables.2. Choosing question format.3. Choosing response modes.4. Preparing questions/items5. Construction of the instrument.6. Pilot testing ² Test for reliability and validity.7. Make required adjustments.8. Repeat 6.