2. Pilot studies. Pilot studies collect data from ultimate subject of the research project to serve as a guide for the larger study. The data collection methods are informal, and the findings may lack precision because rigorous standard are relaxed. 3. Case studies A technique that intensively investigates one or a few situations similar to the researcher’s problem situation. 4. Experience surveys A technique in which individuals who are knowledgeable about a particular research problem are surveyed. In attempting to understand the problem at hand, managers may discuss issues and ideas with top executives and knowledgeable managers who have had personal experience in the field. Although, theoretically, a research study can be classified in one of the above objectives–perspective categories, in practice, most studies are a combination of the first three; that is, they contain elements of descriptive, correlational and explanatory research. 50
4.3 Types of research: mode of enquiry perspective The third perspective in our typology of research concerns the process you adopt to find answers to your research questions. Broadly, there are two approaches to enquiry: 1. the structured approach. 2. the unstructured approach. In the structured approach everything that forms the research process – objectives, design, sample, and the questions that you plan to ask of respondents – is predetermined. The unstructured approach, by contrast, allows flexibility in all these aspects of the process. The structured approach is more appropriate to determine the extent of a problem, issue or phenomenon, whereas the unstructured approach is predominantly used to explore its nature, in other words, variation/diversity per se in a phenomenon, issue, problem or attitude towards an issue. For example, if you want to research the different perspectives of an issue, the problems experienced by people living in a community or the different views people hold towards an issue, then these are better explored using unstructured enquiries. On the other hand, to find out how many people have a particular perspective, how many people have a particular problem, or how many people hold a particular view, you need to have a structured approach to enquiry. Before undertaking a structured enquiry, in the author’s opinion, an unstructured enquiry must be undertaken to ascertain the diversity in a phenomenon which can then be quantified through the structured enquiry. 51
4.3.1 Experimental Research This type of research follows the principles found in natural sciences. Experimental manipulate sets of conditions either in an artificial laboratory setting or in real life and then measure differences in the way people respond in those situations. Experiments are ideal for explanatory research because they can directly address cause-and-effect relationship issues. Comparison of research design Exploratory Descripti Causal ve To determine the Objective To discover To describe cause-and-effect relationship the insights the problem between the independent and and new ideas at hand and dependent variables. about the relate to the Manipulation of problem at characteristic independent variables and hand. To s of the measure the effect on the confirm a population dependent variables. problem. under study. Laboratory Characte Flexible, Based on experiments, changes in the ristics versatile, and specific level of independent does not objectives, variables, changes in the government based on a research policy on import proper questions, tax etc. design and research hypotheses of Method Expert the study Structured interviews, data pilot collection surveys, method, analyze proper secondary statistical data and analysis conduct procedure qualitative research 52
4.3.2 Quantitative & Qualitative Research Quantitative research involves the collecting of data in the form of numbers. Quantitative data are usually associated with empirical social scientific approaches to measurement. The principal methodological technique here comprises surveys and experiments. Qualitative research involves collecting data in the form of words and images. Qualitative research comprises methodologies more often preferred by interpretive and critical school of social science that emphasize interpretation over numerical measurement. 4.4 The Timing of Research 4.4.1 Cross-sectional research Cross sectional design refers to the study in which data collection is gathered from the sample only once. This may involve the analysis of many different groups of people at one point in time or an in-depth examination of one group or organization. 4.4.2 Longitudinal research This research collects data at more than one point in time. Or sometimes in longitudinal design, the data is collected from the same sample on the same variables at two different times. It may consist a series of cross-sectional surveys conducted with different people but asking the same or similar questions, or it may be conducted with the same people who are repeat- interviewed or surveyed. 53
There are three distinct type of longitudinal research: 1. Panel study Involves the observation of or data collection from same people on two or more occasions. 2. Time series analysis Involves collecting data using the same measures at different point in time, but not necessarily with the same people. 3. Cohort analysis A longitudinal study that examines cohort. A cohort is a set of individuals (or groups, organizations) that share a similar experience in a specified time period. 54
5 SAMPLING DESIGN What you’ll learn in this chapter: • What is sampling design • How to choose sampling design 5.1 Selecting the Sampling Design This involves selecting a relatively small number of elements (sample) from a larger defined group (population) and expecting the information gathered from the small group will enable judgments about the larger group. Characteristics of a good sample Accessible Low cost Sample Sample size For size mula sampling Experience from previous Non researchers sampling Depend on judgement 55
Important qualitative factors in determining the sample size 3. The importance of the decision 4. The nature of the research 5. The number of variables 6. The nature of the analysis 7. Sample sizes used in similar studies the major steps in sampling include: Defining the Determi Determining population ning the the sampling Conduct field sample frame design work Select Determining actual sample size sampling units 56
5.2 Choose Point in Sampling Design Is representiveness of sample critical for the study? es o Choose one of the General Assessing non-probability izibility differential sampling design parameters in Choose Choose Choose subgroups of If the purpose of simple systematic cluster population study mainly is: random sampling sampling sampling if not all enough subgroups budget es o To obtain To obtain C Choose quick information relevant dispropor even if hoose tionate to and available proportio stratified unreliable only with certain random informatio nate sampling groups stratified n random sampling Choose convenien Looking Need c for response of e special interest pl s informatio minority groups? n that only Choose few quota experts sampling can provide? Choose judgmental sampling 5.3 Type of Sampling Technique 57
5.3.1 Probability sampling Techniques 8. In the probability sampling, any element of the population has an equal chance or equal probability of being selected as a sample. 9. It is also called random sampling 10. Ensures information is obtained from a representative sample of the population 11. Therefore, the results of the research can be used in generalizations pertaining to the entire population. Type of When to Use Advantages Disadvanta Sampling It ges Simple Time random When the Ensures a high consuming sampling population degree of and tedious members are representatives Systematic similar to one Less random sampling another on Ensures a high than simple important degree of random Stratified variables representatives sampling random and no need to sampling When the use a table of Time population random consuming members are numbers and tedious similar to one another on Ensure a high important degree of variables representatives of all the strata When the or layers in the population is population heterogeneou s and contains several different groups, some of which are 58
related to the topic of study Cluster When the Easy and Possibly sampling population convenient members of consists of units are units rather different than from one individuals another, decreasing the technique’s effectivenes s 5.3.2 Non-probability Sampling Strategies 12. Non-probability sampling is a sampling technique where the samples are gathered in a process that does not give all the individuals in the population equal chances of being selected. 13. Subjects in a non-probability sample are usually selected on the basis of their accessibility or by the purposive personal judgment of the researcher. 14. Therefore, the results of the research cannot be used in generalizations pertaining to the entire population. 59
Type of When to Use It Advantages Disadvantages Sampling Convenience When the Convenient and Degree of sampling members of the inexpensive generalizability is population are questionable Quota convenient to Ensures some sampling sample degree of Degree of representativen generalizability is When strata are ess of all the questionable present and strata in the stratified population sampling is not possible 5.3.3 Differences between probability and non-probability sampling techniques Comparison Factors Probability Non- sampling probability sampling List of the Population Complete list Not necessary Elements necessary Time Requirement Time consuming Low time consumption Estimates of Population Unbiased Parameters Biased Measurement of Sampling Statistical Error measures No true measure Sample Good, assured available Representativeness Suspect, undeterminable 60
Probability Non-probability Allows use of statistics, Exploratory research, generates tests hypotheses hypotheses Can estimate population parameters Population parameters are not of interest Eliminate bias Adequacy of the sample can’t be Must have random known selection of units Cheaper, easier, quick to carry out 5.3.4 When to use non-probability sampling 61
5.4 Sample Size Factors to determine sample size: 1. The variability of the population characteristics under investigation. The greater the variability of the characteristics, the larger the size of the sample necessary for that research. 2. The degree of precision desired in estimating the population characteristics. The more the precise the result from the study required, the larger the sample size needed to be drawn from the population. Krejcie and Morgan (1970) greatly simplified the size decision by providing a table that ensures a good decision model. The interested student is advised to read Krejcie and Morgan (1970) as well as Cohen (1969) for decisions on sample size. 62
6 ERRORS IN SURVEY RESEARCH What you’ll learn in this chapter: • Types of error in survey research • How to check the goodness of data 6.1 Total Survey Error Errors can occur during the implementation of a survey research. In evaluating the quality of a survey research, the researcher must estimate the accuracy of the survey by identifying the two major sources of errors: 1. Random sampling error 2. Non sampling error 6.2 Random Sampling Error Definition Sampling error is any type of bias that results from mistakes in either the selection process for prospective sampling units or in determining the sample size The difference between the result of a sample and the result of a census conducted using identical procedures. Random sampling error occurs because of chance variation in specific selection of sampling units. The sampling units, even though properly selected according to sampling theory, may not perfectly represent the population, but they are generally reliable estimates. If there is slightly difference between the true population value and the sample value, hence a random sampling error will occur. 63
How to minimize the sampling error Sampling technique • Using appropriate sampling technique when drawing the sample from its population Size of • Increasing the size of sample or respondents to sample be obtained from its population 6.3 Non Sampling Error (Systematic Error) Definition Non Sampling error is bias that occurs in a research study regardless of whether a sample or census is used; e.g., bias caused by measurement errors, response errors, coding errors, etc. Systematic error results from some imperfect aspect of the research design or from a mistake in the execution of the research. These are errors that come from such sources as sample bias, mistakes in recording responses, and non- responses from persons not contacted or refusing to participate. Systematic error is divided into respondent error and administrative error. Respondent error is further divided into non- response error and response bias. 64
6.3.1 Selection bias: Misspecifying the target population. Undercoverage: Falling to include all of the target population in the sampling frame. Overcoverage: Include population units in the sampling frame tnhoattianrtehe target population. Substituting a convenient member of a population for a designated member who is not readily available. Falling to obtain responses from all of the chosen sample. How to eliminate sample bias? Good research design Drawing samples in an unbiased way Approach to deal with nonresponse Prevent it. Design the survey Take a representative subsample of so that nonresponse is low. This is the nonrespondents; use that by far the best method. subsamples to make inferences about the other nonrespondents. Use a model to predict values for the nonrespondents. Weighting class adjustment methods implicitly use a model to adjust for unit nonresponse. Imputation Ignore the nonresponse often adjusts for item nonresponse, and parametric models may be used for either type of nonresponse. 65
6.3.2 Respondent error It is the statistical difference between a survey that includes only those who responded and a survey that also includes those who failed to respond. A non-response occurs if someone is not at home and a subsequent call back also finds the subject not at home. 6.3.3 Response bias Further divided into 2, that is deliberate falsification and unconscious misrepresentation. 1. Acquiescence bias A category of response bias in which individuals have tendency to agree with all questions or to indicate a positive connotation. 2. Extremity bias A category of response bias that results from response styles varying from person to person, some individuals tend to use extremes when responding to questions. 3. Interviewer bias Bias in the response of subjects due to influence of the interviewer 4. Auspices bias Bias in the responses of subjects caused by the respondents being influenced by the organization conducting the survey. 5. Social desirability bias Bias in the responses of subjects caused by respondents’ desire, either consciously or unconsciously, to gain prestige or to appear in a different social role. 66
6.3.4 Administrative error It is error caused by the improper administration or execution of a research task. The errors include: 4. Data processing error 5. Sample selection error 6. Interviewer error 7. Interviewer cheating 6.3.4.1 Approach to reduce measurement error • Test your questions • Interview potential respondents to see if they interpretthe questions as you intend. • Write clear questions. • If a respondent does not know how to answer a question, the interviewer is likely to have more influence on the response in a self administered survey, unclear questions can lead to more variability or bias in the responses. Open- ended questions may be more susceptible to interviewer effects than closed questions. • Provide training and supervision for interviewers so they act consistently • Interviewers should read the questions exactly as written, and should not indicate that oneresponse is preferred over another. 67
6.3.5 Testing Goodness of Data Definition Reliability occurs when a test measures the same thing more than once and results in the same outcomes. 6.3.5.1 Reliability Reliability consists of both an observed score and a true score component. Observed score is the score you actually record or observe. Meanwhile the true score is a perfect reflection of the true value of a variable, given no other internal or external influences. The measurement process and the theory of reliability always assume a true score is there. Therefore, the error score is all those factors that cause the true score andthe observed score to differ. 68
Methods to increase reliability: 1. Increase the number of items or observation. 3. Eliminate items that are unclear. 4. Moderate the degree of difficulty of the tests. 5. minimize the effects of external events. 6. Standardize the conditions under which the test is taken. 7. Standardize the instruction. 8. Maintain consistent scoring procedures. Types of reliability: 1. Stability of measures The ability of a measure to remain the same over the time. i. Test-retest The reliability coefficient obtained by repetition of the same measure on a second occasion. ii. Parallel form When responses on two comparable sets measures tapping the same construct are highly correlated. 2. Internal consistency of measures The indicative of the homogeneity of the items in the measure that tap the construct. i. Interim A test of the consistency of respondents’ consistency answers to all the items in a measure. ii. Split-half Reflects the correlations between two halves of an instrument. The values vary depending on how the items in the measure are split into two halves. 6.3.5.2 Validity Definition Validity is a test on how well an instrument that is developed measures the particular concept it is intended to measure. 69
Three aspects of validity: 1. Validity refers to the results of a test, not the test itself. 2. The results of a test are not just valid or invalid. This progression occurs in degrees from low validity to high validity. 3. The validity of results of a test must be interpreted within the context in which the test occurs. Validity Description Content validity Ensures that the measures include and adequate and Criterion- related representative set of items that tap the concept. Established when the measures differentiate individuals Construct on a criterion it is expected to predict. Concurrent Established when the scale discriminates individuals who are known to be different; that they should score differently on the instrument. Predictive Indicates the ability of measuring instrument to differentiate among individuals with reference to a future criterion. Testifies how well the results obtained from the use of the measure fit the theories around which the test is designed. Convergent Established when the scores obtained with two different instruments measuring the same concept are highly correlated. Discriminant Established when based on theory, two variables are predicted to be uncorrelated, thus the scores obtained by measuring them are indeed empirically found be so. The relationship between reliability and validity: A test can be reliable but not valid, but a test cannot be valid without first being reliable. 70
7 SOURCES OF DATA What you’ll learn in this chapter: • Types of data sources • Method of collecting data 7.1 Type of Data Sources Secondary and primary data sources The sources of data and information in research can be classified into primary data and secondary data. The primary data is basically the data that is collected and assembled by the researchers of a current research problem. Some experts refer to primary data as “first-hand” data structures because the researcher collect the data on their own. The secondary data includes historical information or data structure which was previously collected and assembled for solving other problems and not the current problem. These data can be in form of newspaper report, annual report of certain department, or the published report by certain organization. 7.2 Method of Collecting Data Data collection methods can be categorized into five methods that are commonly used by the researcher to get the primary data. Table 1.5 gives the description for every method of data collection. 71
Data Collection Table 7.1 Methods of data collection method Direct Interview/ Description Personal Interview Trained interviewer will get the information from Telephone respondents depending on the objective of the survey Interview and based on questionnaires. It is commonly used in Mailed marketing research and social survey. Questionnaire It is possible to use telephone interview as a Direct personal interview. It is conducted by telephone Observation services line and respondents are asked to respond to the questionnaire. Online Surveys Questionnaires are mailed to the respondent included from the sample. A carefully constructed questionnaire will encourage participation by the respondents. It does not require interviewers. Some questionnaires do enjoy high initial response rates. One example is questions asked on warranty cards that must be returned to the manufacturer for warranty coverage of a new product. The response rate to government questionnaires tends to be much greater than those distributed by private sector. Questionnaires that are distributed to the members of an affiliation/association also enjoy high response rate since they have a common interest among the membership. Observation is used in work-study and in organization. A quality control exercise in a factory making light bulbs in which the data are collected by taking bulbs from the production line and observing the number of hours they last. Used by social scientists in learning the habits and customs of communities. Nowadays, online questionnaires are a very popular method to collect data. It is a cheap way to reach respondents globally. Response rates are higher than mailed questionnaire method. Nevertheless, respondents have to be reminded through emails. 72
There are various methods of data collection but not all can be used in a research. It depends on the characteristics of research we focus on. For example, we cannot observe motivation of an employee to do an assignment. It is impossible to measure motivation based on observation. Some of the advantages and disadvantages of the methods are summarize in Table 7.2 below. 73
Table 7.2 Advantages and disadvantages for each methods of data collection Data Advantages Disadvantages Collection method Direct Interview/ § Obtained higher percentage of response than other methods. § The cost is high (pay interviewers salary, traveling etc). Personal § There are many respondents, which are co-operative than § Interviewer must be trained so that the data received is accurate Interview non- cooperative. and doesnot show biasness. Telephone § Interviewer can clarify any terms that are not understood § Error may occur in recording statements/ response. Interview § Interviewer should be supervised to ensure that he has actually by the respondents. Mailed § Interviewer can detect whether the respondent is giving an contacted the proper respondent, and that his behavior was Questionnaire appropriate so that he did notfill in the questionnaire himself. accurate answer from their characteristics. § The interviewer can ask a lot of questions. This approach can be time consuming. § The interviewer can note specific reactions and the § Limitation is that we restrict ourselves to only among the telephone environment surrounding the respondent. users. § Lesser time is needed to obtain the information. § The investigator can also monitor the interviews to follow the § Only short questions can be asked. § The response rate is lower than the personal interview. specified procedure. § Interviewer has to find the correct time to interview the respondents. § Less expensive than the personal interview owing to the § The lowest rate of response, sometimes less than 20%, since we eliminationof the traveling expenses. have the least contact with the respondents. It is always necessary to contact non-respondents to the first mailing through follow-up § Convenience since wider region can be covered. letters. Free gifts and financial incentives may increase the rate a § Entails a relatively lower cost. little bit. § Information received from the respondents is free from the § Respondents may not return the questionnaire within the specified biasness of the investigator. time. § No prior arrangements necessary to conduct the survey. § Information may be elicited or sensitive which would be § Respondents might not understand the question asked. If they were helpedto answer the questionnaire, this may result in non- embarrassingto obtain in face-to-face interview. representative sample. § Enough time is provided to the respondents to § No one can assist the respondents in answering the questionnaire. complete the questionnaires. Therefore, the respondent may become frustrated and not bother completing orreturning the questionnaire. Observation § The observer needs to be highly skilled and unbiased. § It records what actually happened rather than what people say § The observer will not get an accurate data if the respondents know would happen. that they are being observed. § The data are not affected by the respondents. § Present observation tells us nothing about past or future happenings. 74
7.3 Secondary Data Secondary data refers to historical data structures of variables that have been previously collected and assembled for some research problem or opportunity situation other than thecurrent situation. This includes data on financial markets in KLSE Industrial, data on company directors’ remuneration and population demographic data. Definition Secondary information refers to information (facts or estimates) that has already been collected, assembled and interpreted at least once. 7.3.1 Sources of secondary data The major sources of secondary data are as discussed below. 7.3.1.1 Internal Sources Internal organization sources vary depend on the usage. For example, the departmental reports, production summary, financial and accounting reports etc. 7.3.1.2 External Sources External data are created, recorded, or generated by entity other than researchers’ organization. Published data can be classified into five categories: 1. Computerized databases 2. Books 3. Periodicals 4. Government sources such policies through their respective departments. 5. Miscellaneous sources such master’s theses, dissertations and research records. 75
7.3.1.3 Procedure for collection of secondary data Researcher should ensure the appropriate data source and organize data into appropriate form. The task involved in searching secondary data are through: 1. The bibliography search 2. The computerized search The advantages of computer search are the speed, flexibility and up to date information. Meanwhile, the disadvantages are imprecision, wide capability and limited resources (for journal prior mid 1970). 3. The library search – Books, periodicals, government sources, miscellaneous sources. 76
8 SURVEY INSTRUMENT What you’ll learn in this chapter: • Characteristics of a good questionnaire • Types of rating scales with examples. The survey instrument may be a questionnaire or an interview schedule. In a self-administrated questionnaire, respondents read the instructions and write or mark their answer to the questions. Questionnaires begin with an introduction, which explain the purpose of the survey and gives instructions to the respondent. The questions fall into two main categories: demographic and content questions. Demographic questions seek descriptive information about the respondents such as age, gender, occupation, marital status and so on. Most items on questionnaire are content items, which ask about opinion and attitude such how they feel about environmental issues, or any issues are subjective and vary among individuals. There is no right and wrong answer. The steps in survey research: 1. Determine what area of information to be sought (Basedon objectives of the study) 2. Define the population to be studied. 3. Decide how the survey is to be administered. 4. Construct the first draft of the survey instrument, edit and refine the draft. 5. Pretest the survey with a subsample, refine it further. 6. Draw a representative sample. 7. Administer the final form of the instrument to the sample. 8. Analyze, interpret and communicate the result. 77
8.1 DESIGNING A QUESTIONNAIRE Definition Questionnaire is a written instrument that contains a series of questions or items that attempt to collect information on a particular subject. Questionnaires may be handed out personally by the researcher to his respondents or may be posted the mail. The steps in questionnaire development: 1. Identify information required 2. Determine method of data collection 3. Finalise the content of question 4. Determine the wordings of question 5. Determine the structure of responses 6. Decide on the order of question 7. Decide on the form and layout 8. Pre-test the questionnaire. The researchers should construct a questionnaire to encourage participation from the respondents. Several steps to be considered when designing a questionnaire are listed below: 1. Specify the information to be collected. The source of information should be from the objectives and specific aims of the survey. Therefore, the purpose of the survey should always keep in mind. 2. Questionnaire should be as short as possible. A long list of questions may lead boredom and incorrect answers. 3. Avoid ambiguous question. Questions must be kept simple and phrased to imply the same meaning to all respondents. Two simple questions are better than a single complicated one. 4. Avoid leading questions such as: ‘Don’t you think the courts are too harsh with drug trafficker?’. This questionsuggests 78
the answer the interviewer wants to hear, and the respondents may agree with the interviewer because that is the easiest response. 5. Questions should be organized systematically and should have some kind of natural logical sequence that the respondent can appreciate. This will help to sustain interest. 6. Avoid questions that need calculation. 7. Avoid questions on sensitive issue or may be confidential to the respondents. 8. Avoid double barreled questions. The items in the questionnaire also can take several forms other than previously discussed rating scales such as open-ended items, multiple choice items. 8.1.1 Open-ended question: a. Respondents have the freedom to provide their views or opinions concerning certain issues. b. Respondents answer the questions in their own word. Example: Which Hotel would be your next preferred choice to stay aside from this Hotel? Please fill in the name of the hotel 8.1.2 Close-ended question: a. The responses are already provided in the questionnaire. b. The respondents have the option to choose “others”if none of the choices are related to them. 79
Example: Did you experience any problem which is of concern to you during your stay with us? Yes No 8.1.1 Rating scales The following rating scales are often used in organizational research: Rating Scales Example Do you own a car? Dichotomous scale Yes No Category Scale Where in London do you reside? East London South London West London North London Outskirts Semantic Responsive _______ Unresponsive Differential Beautiful _________ Ugly scale Courageous _______ Timid Numerical How pleased are you with your new state scale agent? Extremely Pleased 7 6 5 4 3 2 1 Extremely Displeased 80
Itemized rating scale Respond to each item using the scale below and indicate your response number on the line by each item. 1 2 34 5 Very Very unlikely Unlikely Neither Likely likely 1 Unlikely or 2 likely 3 I will be changing my job within next 12 months. I will take on new assignments in the near future. It is possible that I will be out of this organization within the next 12 months Likert Scale/Summated scale The Likert Scale is designed to examine how strongly subjects agree or disagree with statements on five-point scale with the following anchors: Strongly Disagree Neither Agree Strongl Disagree 2 Agree Nor 4 y agree Disagree 1 5 3 My work is very interesting 12 34 5 I am not engrossed in my work all 12 34 5 day 12 34 5 Life without my work would be dull Fixed or constant sum scale In choosing a toilet soap, indicate the importance you attach to each of the following five aspects by allotting points for each to total 100 in all. 81
Example: Fragrance 100 Color Shape Size Texture of lather Total points 82
9 DATA MANAGEMENT What you’ll learn in this chapter: • Stages in data preparation • How to handle missing data. 9.1 Preparation Stages 9.1.1 Data coding Refers to the process of identifying and classifying each response with a numerical score or symbol. The researcher must specify the column position, number of variables, name of variable and coding instructions. 9.1.2 Data entry Includes tasks involved with the direct input of the coded data into some specified software package that ultimately allows the research to manipulate and transform the raw data into useful information. 9.1.3 Data Editing The process in which raw data are screened for errors that might occur during sampling or data collection process. The purpose of data editing is to increase the accuracy and precision of the data by checking certain omissions such as missing data. Editing of a missing data or no response. 1. Go back to the respondent if the number of missing responses on that particular respondent is large and the number of sample in the study is small. 83
2. Assigning missing values if the number of the respondents involved in missing data is small. The proportion of missing data for each respondent is smalland if the missing data are not the main variables in the study. 3. Excluding missing responses if the proportion of respondents with missing data is small while the sample size is large. The proportion of missing data is large and impossible to edit where the missing data involved is the main variables in the study. You can also delete it if the respondent with missing data has similar characteristics with other respondents in the study. Step in data processing (Kumar, 2014) are as follow: 84
85
10 DATA ANALYSIS What you’ll learn in this chapter: • Types of data analysis • Choosing data analysis according to scales of measurement Statistics can be divided into two, descriptive statistics and inferential statistics. Type of analysis will be used in your study are depend on the objectives and type of variable u chose. 86
10.1 Descriptive Statistics Researchers go to great lengths to obtain the central tendency, the range, the dispersion, and other statistics for every single item measuring the dependent and independent variables. Descriptive statistics for a single variable are provided by frequencies, measures of central tendency and dispersion. These are now described: Scale Measures Measures Visual Measure of Visual of central of Summary relation summary of tendency dispersion (between relation (for a (for a variables) (between single single variables) variable) variable) Nominal Mode - Bar chart, Contingency Stacked Ordinal pie chart table (cross bars, Interval Median Semi- tab) Clustered Interquartile Bar chart, bars Ratio range pie chart Contingency Stacked table (cross bars, Arithmetic Min, Max, Histogram, tab) Clustered mean standard scatter plot, bars deviation, box- and- Correlations variance, whisker Scatterplots coefficient plot Correlations Scatterplots of variation Histogram, scatter plot, Arithemetic Min, Max, box- and- or geometric standard whisker mean deviation, plot variance, coefficient of variation Types of tables Depending upon the number of variables about which information is displayed, tables can be categorised as: • univariate (also known as frequency tables) – containing information about one variable, for example Table 16.1 87
• bivariate (also known as cross-tabulations) – containing information about two variables, for example Table 16.3 • polyvariate or multivariate – containing information about more than two variables, for example Table 16.4. 88
10.2 Inferential Statistics Test for significant relationship based on the scales of measurement are as follow: Nominal versus nominal -Chi square test -Cramer’s V Test Phi- Coeficient Test for significant relationship Ordinal versus Ordinal Interval/ratio versus Interval/ratio -Chi square test -Pearson’s -Spearman Rank -Kendall’s Tau Next, list of all possible data analysis is mapped for reference. 89
90
Example data analysis based on research hypothesis: Parametric Test Nonparametric test One sample t-test Research Hypothesis One sample Sign Wilcoxon Sign-rank RH: Patient not satisfied with the service provided by ABC Median test RH: There is significant difference in satisfaction level towards the Independent t-test Mann Whitney test service provided at ABC between male and female patients. RH: There is significant difference in patient’s perception towards the Paired t-test Sign test for two related service provided at ABC before and after they recieve the treatment. samples RH: There is significant difference in satisfaction level towards the ANOVA Wilcoxon mathed pairs service provided at ABC among races. sign rank test RH: There are significant difference in effectiveness of the fourtreatment. Repeated Extension of the RH: Patient’s satisfaction significantly effect patient’s loyalty. Measure median test Kruskal Wallis test ANOVA Friedman Test Pearson correlation Spearman Rank correlation 91
11 AN OVERVIEW IN REPORT WRITING What you’ll learn in this chapter: • Chapters in report writing • Do and don’t in report writing. Chapter Chapter Title Contents Chapter Introduction Background to the topic/are. The One problem/issue that wasunknown. Literature Review What the researcher wanted to Chapter Methodology investigate Two Results Chapter What was already known about Three the topic/area Chapter How the researcher went about Four looking for answer Chapter What the researcher found Five Discussion/Concl The implications or significance u sion of these findings A thesis is written in an academic style, which is different from the style of other types of writing. Furthermore, it is not a style that is natural to most writers. Some of the features that characterize this style includes: 1. An impersonal way of writing • In academic writing, the people who are involved in the writing and reading of a thesis (reader or writer) are not usually mentioned directly in the text. So, the first person and second-person pronouns ‘I’ or ‘you’ are not used. • Avoid asking questions in your writing • Avoid giving commands. 92
Eg: ‘Refer to Figure 6 for a breakdown of the data’. Instead write, ‘Figure 6 shows a breakdown of the data’. 2. Use of formal vocabulary o Avoid slang o Avoid shortened words (eg: Isn’t, didn’t etc.) o Avoid phrasal words Eg: Phrasal verbs Single word verbs Look at Investigate, examine Help out assist, aid Set up establish, install Get rid of eliminate, expel 2. Objectivity or neutrality in expressing ideas o Being objective (or neutral) means that you do not let your personal feelings or opinions influence the manner in which you write about you research in your thesis. e.g.: The results were disappointing. The results were inconclusive or did not show the expected correlation. (Neutral) 3. Use of passive voice constructions o To avoid any mention of the writer or the reader in the text is to write in the passive voice instead of active voice. e.g.: First, I interviewed 10 key members of the organization. (Active) First, 10 key members of the organization were interviewed. (Passive) *Excessive use of passive voice is now discouraged in most other formal writing (e.g.: in business and professional writing) but still accepted or even expected in thesis writing. 93
4. Use of hedging and tentativeness in expressing ideas o Use of qualified language or hedging refers to ways of expressing ideas tentatively, without sounding over-confident or making extravagant claims that may be difficult to justify. e.g.: “appears”, “suggest”, “indicates” etc. 5. Use of longer sentences which combine several pieces of information 6. A system for citing sources of information 4 main reasons for citing: a. To help readers to find information you have referred to b. To tell the readers where an idea comes from c. To show you have read and understood the research published in your area. d. To avoid plagiarizing. Citation style: - APA (American Psychological Society) - Harvard system - IEEE and etc. 7. Formality in use of conventions relating to contractions, numbers, abbreviation. Format : Times New Roman (12) or Arial (11) Font Size : 1.5 Spacing : Position (Bottom of Page) Page Number : Centre Alignment 94
Guidelines for STA220 report CHAPTER REQUIREMENT MARKS 5 marks Acknowledgement 10 marks Table of contents List of tables 10 marks List of figures 20 marks CHAPTER 1 1 paragraph 20 marks 10 marks 1.1 Introduction At least three 95 1.2 Problem Statement objectives 1.3 Research Objectives Depend on objectives 1.4 Research Question Depend on research 1.5 Research Hypotheses questions 1.6 Significance of the study 1.7 Scope of the study 1.8 Limitation of the study Chapter 2 : LITERATURE At least 10 citations REVIEW 2.1 Literature Review CHAPTER 3 : At least 30 samples RESEARCH METHODOLOGY 3.1 Study Population 3.2 Study Design 3.3 Description of sample Questionnaires 3.3.1 Included criteria Probability 3.3.2 Excluded criteria sampling technique only 3.4 Measuring instrument At least 30 samples 3.5 Method of Data Collection At least 10 samples 3.6 Sampling technique 3.7 Sample size 3.8 Pilot study 3.9 Theoretical framework 3.10 Procedure for data analysis 3.10.1 Descriptive Statistics 3.10.2 Inferential statistics 3.11 Scope and limitations of the study Chapter 4 : RESULTS AND Descriptive DATA ANALYSIS statistics t- test, chi 4.1 Descriptive Statistics square, ANOVA, multiple regression 4.2 Data Exploration (extra marks will be 4.2.1 Continuous Variable given) 4.2.2 Categorical Variable CHAPTER 5 : CONCLUSION At least 3 recommendations AND RECOMMENDATION 5.1 Conclusions 5.2 Recommendations References At least 10 references Appendices Additional outputs from SPSS
12 REFERENCES Awang, Z. (2012). Research Methodology and Data Analysis (2nd Ed.). Penerbit PRESS UiTM. Blaxter, l., Hughes, C. & Tight, M. (2010). How to research (4th Ed.). McGraw Hill. Fink, A. (2005). Conducting research literature reviews: from the internet to paper (2nd Ed.). Sage Publications. Graziano, A. M. & Raulin, M.L. (2010). Research methods: a process of inquiry (7th Ed.). Allyn & Bacon. Jackson, S.L. (2011). Research Methods: a modular approach (2nd Ed.). Wadsworth Cengage Learning. Mohd Noor, N. & Page, G.M. (2010). Writing your thesis. Prentice Hall. Mahmud, Z. (2008). Handbook of Research Methodology: A simplified version. Penerbit PRESS UiTM Kumar, R. 2014, Research Methodology, 4 Ed., SAGE Publications Limited [ISBN: 9781446269978] Salkind, J. N. (2012). Exploring Research (8th Ed.). Pearson Education. Sekaran, U. & Bougie, R. (2010). Research Methods for Business (5th ed.). John Wiley & Sons. 96
Search
