Section 2 How to Design Quantitative Research Elements

Section 2: How to Design Quantitative Research? Elements of Research Design Formulation of Research Question and Hypotheses From Conceptualization to Measurement 2022/1/3 2015 Research Methods (HK) 1

Elements of Research Design Purposes of Research Unit of Analysis Time Dimension Theory, concept and variable 2022/1/3 2015 Research Methods (HK) 2

Purposes of Quantitative Research Exploratory Studies Three general purposes of exploratory studies: To satisfy the researcher’s curiosity and desire for better understanding To test the feasibility of conducting a more complex research To develop the methods to be employed in a more careful study Descriptive Studies The major tasks of descriptive study are to discover social facts and to describe social reality , e. g. U. S. Census and HK Census The findings of descriptive studies are, for example: the percentage of smoking female; the age mean of the students studying overseas… Explanatory Studies It should include three elements: A behaviour or an attitude (such as racial prejudice) Determinant of such behaviour or attitude (e. g. educational level, sex, religion…. ) Reason why the causal element influences the behaviour or attitude (e. g. why men have greater degree of racial prejudice? …) Explanatory studies usually seek to ask ‘why’ and ‘how’ questions. 2022/1/3 2015 Research Methods (HK) 3

Unit of Analysis Definition The wide range of variations in what or who is studied, described analyzed. Types of Unit of Analysis Individuals Social Groups Organization Social Artifacts Related concept Ecological fallacy 2022/1/3 2015 Research Methods (HK) 4

Ecological Fallacy “It arises from a mismatch of units of analysis. It refers to a poor fit between the unit for which researcher has empirical evidence and the units for which he or she wants to make statements. ” Neumen 2002 2022/1/3 2015 Research Methods (HK) 5

Example for ecological fallacy n Candidate: n Ms. L n Electional District: n Shatin n Estate: n XX n Block: n A, B, C and D n Average Age of Voters: n n n Obtained Result: Block A— 40, Block B— 38, n Block C— 35, Block D— 33 n Block C and Block D gave a large proportion of their votes than the others. Possible Conclusion: Younger voters were more likely to vote for Ms. L than older voters. 2022/1/3 2015 Research Methods (HK) 6

…Example for ecological fallacy Committed the ecological fallacy: It may have been the older voters in those young blocks who voted for Ms. L. Problem: We had examined blocks as our units of analysis and wish to draw conclusions about voters. Read Neumen 2003 (p. 158) for the ecological fallacy made by Durkheim 2022/1/3 2015 Research Methods (HK) 7

Theory, Concepts & Variable Relationship between Theory Concept Variables Theory A theory is a systematic explanation for the observed facts & laws that relate to a particular aspect of life: deviance, social learning, selfconcept, cultural capital…. . 2022/1/3 2015 Research Methods (HK) 8

Concepts referred as “the basic blocks of building theory”. Babbie E. --The concepts relevant to theory of cultural capital would include socio-economic status, social network, satisfaction gained from work, educational level, resources of obtaining career or education related information…… 2022/1/3 2015 Research Methods (HK) 9

Variable Ø The root of the word comes from ‘vary’, meaning--something varies. Ø A variable has values which can change. Ø Variables are a matter of observation and measurement. Ø They need to be more specific than concepts. 2022/1/3 2015 Research Methods (HK) 10

Example to show those terms are being used in a real research situation THEORY Theory of Cultural Capital is used to explain why middleclass children achieve better academic attainment. Nowadays it is generally believed that money can buy education, that is to say economically better off children have a greater chance of enjoying quality education. Differences (variation) in education of parents, living styles, habitus, norms and values in education, and social network may well be the explanation for the generally better/worse academic performance. All those are not directly related to economy, some of them are social and some cultural! 2022/1/3 2015 Research Methods (HK) 11

Concepts & Variable with examples Research problems are conveyed with a set of concepts. (e. g. …. . ) Concepts are abstractions representing empirical phenomena. Moving from conceptual to empirical level, concepts are converted into variables which eventually appear in hypotheses and are tested. 2022/1/3 2015 Research Methods (HK) 12

Concepts & Variable with examples (Con’t) Concept: Education Variable Values Year of schooling State actual years Type of educational qualification obtained Secondary, Matriculated, University or Above university Educational level F. 1, F. 2, F. 3……. 2022/1/3 2015 Research Methods (HK) 13

Types of variables Explanatory Variables Dependent variable (DV) & Independent variable (IV) Extraneous Variables Antecedent variable & Intervening variable 2022/1/3 2015 Research Methods (HK) 14

Explanatory Variables with examples Independent variable (predicted variable) are presumed to cause or determine a dependent variable Dependent variable are assumed to depend on or be caused by another Example n One’s parents’ education level (IV) may determine his/her academic performance (DV) n One’s academic performance (IV) may determine his/her social network (DV) Some variables can be an independent and a dependent variable in different cases. 2022/1/3 2015 Research Methods (HK) 15

Extraneous Variables Extraneous variable are NOT part of the explanatory set. Antecedent variable It occurs prior in time to both the IV & DV. Parents’ income, students’ gender and age, each of them will affect both learning motivation and academic attainment. Intervening variable A variable is intervening if it is an effect of IV and a cause of the DV. They are those not easy, or even impossible, to observe and also difficult to predict 2022/1/3 2015 Research Methods (HK) 16

Differences between Quantitative and Qualitative variables a. A variable is quantitative if it values consist of numbers and if differences between its categories can be expressed numerically. n 2022/1/3 e. g. I. sense of belonging ranged from 0 -10 b. Qualitative variables have discrete categories, usually designated by words or labels. n e. g. I. male or female n e. g. II. married or never married 2015 Research Methods (HK) 17

Formulation of Research Question and Hypotheses How to Formulate Research Question? What are Testable Hypotheses? 2022/1/3 2015 Research Methods (HK) 18

Research Questions & Research Hypotheses It is difficult to move from a wide topic to hypothesis, but the path from a well-formulated research question to hypotheses is a short one. Research Questions Research Hypotheses 2022/1/3 2015 Research Methods (HK) 19

How to Formulate Research Questions must not be too “grand” or too “broad”. Both experienced and novice researchers can easily be tempted to answer “grand questions”—too broad to be scientifically answerable. 2022/1/3 2015 Research Methods (HK) 20

…Research Questions Researchers should reformulate their problems so that they point to identifiable variables and relationship. A problem statement is needed. It should express or ask a question about a relation between 2 or more variables. That is-n “what causes A? ” e. g. What causes stress? What causes sense of belonging? n “what are the effects of B? ” etc. e. g. What are the effects of political alienation? 2022/1/3 2015 Research Methods (HK) 21

…Research Questions The set of questions to be answered are very important in any research work. Care should be taken not to ask vague questions. They should be more specific. When constructing the research problem, one can start ask “general questions”, then “specific questions”. There are three types of questions for QN research: a. descriptive question b. explanatory question c. exploratory question 2022/1/3 2015 Research Methods (HK) 22

…Research Questions Type of Question Descriptive Question What cultural capitals do Hong Kong middle-class parents have? Explanatory Question i)Do cultural capitals promotes academic attainment? (Trying to explore a relationship between 2 variables) ii)Why are cultural capitals becoming more important in promoting children’s academic attainment? (Trying to identify the factors behind a social phenomenon) Exploratory Question What are the cultural capitals that are beneficial to children’s academic attainment under globalization? (Trying to look into themes when very limited is currently known) 2022/1/3 2015 Research Methods (HK) 23

…Research Questions “Apart from approaching the research from asking questions, there are some studies use hypotheses testing instead of questions asking. ” Read Neumen P. 162 -P. 164 2022/1/3 2015 Research Methods (HK) 24

How to state Testable Hypotheses? What is a hypothesis? Nature of Hypothesis Function of Hypothesis 2022/1/3 2015 Research Methods (HK) 25

What is a hypothesis? The tentative answers to research questions are called hypotheses. Formal definition: a hypothesis is an expected but unconfirmed relationship between 2 or more variables 2022/1/3 2015 Research Methods (HK) 26

Nature of Hypothesis Adequate hypothesis statement about 2 variables should indicate: a. which variable predicts or causes the other n b. how changes in one variable are related to changes in the other n 2022/1/3 2015 Research Methods (HK) 27

Function of Hypothesis To state a specific relationship between phenomena in such a way that the relationship can be empirically tested. Example: Researcher thinks cultural capital generally promotes education attainment. Ø Ø 2022/1/3 An increase in cultural capital will result in an increase in education attainment. Children coming from family with more cultural capital perform better academically. Cultural capital is positively correlated with education attainment. The amount of time used in digital games is negatively correlated with academic attainment. 2015 Research Methods (HK) 28

From Conceptualization to Measurement The steps leading to the measurement of a variable Conceptualization Nominal Definition Operational Definition Measurement Sources of Error 2022/1/3 2015 Research Methods (HK) 29

Conceptualization Concept In our everyday life, we use concepts all the time. But they can be very abstract. They are used to explain or make sense of our experience. Definition of Conceptualization “Conceptualization is a process through which we specify precisely what we will mean when we use particular term. ” Macionis (2003) Baker (1994) calls conceptualization, “…the central intellectual effort required by social research. ” Operationalizing Definition Why defining is necessary? Two types of definitions for concepts 2022/1/3 2015 Research Methods (HK) 30

Why defining is necessary? Giving variables clear and precise definition is vitally important, without it observation can never be carried out effectively. Definitions are used to explain situations, concepts, or phenomena. The most relevant to the research should be applied, by relevant, it is meant the topic and the sample are taken into careful consideration. 2022/1/3 2015 Research Methods (HK) 31

Two types of definitions for concepts Nominal definition Operational definition 2022/1/3 2015 Research Methods (HK) 32

Nominal Definition It is one that is assigned to a term. The specification of nominal definitions focuses our observational tactics, but it does not let us observe. To be able to observe, it is necessary we specify exactly what we are going to observe, how it is done and what interpretations we are going to put on different possible observation. 2022/1/3 2015 Research Methods (HK) 33

Nominal definition: Example (con’t) Concept 1: Socio-economic status (SES) Nominal definition of concept: A composite ranking based on various dimensions of social inequalities n Many aspects in life should be taken into considerations when talking about one’s SES, such as occupational status, income, educational level, money in the bank, property owned, linkage, life styles…… n 2022/1/3 2015 Research Methods (HK) 34

Operational Definition A definition that tells exactly how the concept will be measured. To be more precise, it is a definition that describes the “operations” that are adopted to measure a concept. In the case of the SES, two questions may be asked: What was your total family income during the past 12 months? n What is the highest level of school have you completed? n n Obviously, that’s not sufficient, index and scale will be used to give more valid measurement. (index and scale construction will be discussed at a later stage) 2022/1/3 2015 Research Methods (HK) 35

Levels of Measurement Definition & What are the levels of measurement? Different Levels of Measurement Sources of Error Assessment of Reliability Assessment of Validity 2022/1/3 2015 Research Methods (HK) 36

Definition of Measurement and What are the levels of measurement “. . the assignment of numbers or labels to units of analysis to represent variable categories. ” Babbie (2001) Level of measurement provides a framework for interpreting the categories of a variable. n 4 different empirical rules exit for sorting cases into categories. n Each rule results in different interpretations of the numbers that may be assigned to the various categories. n Various meanings of the numbers are what is meant by levels of measurement. n 2022/1/3 2015 Research Methods (HK) 37

Different Levels of Measurement Nominal Measurement Ordinal Measurement Interval Measurement Ratio Measurement 2022/1/3 2015 Research Methods (HK) 38

Nominal Measurement with examples The lowest measurement. Where cases are classified into 2 or more categories on some variables: gender, material status, race, religious affiliation… A naming or labelling process. Numbers are assigned to the categories as labels or codes for the convenience of collecting and analyzing data. 2022/1/3 2015 Research Methods (HK) 39

Ordinal Measurement with example Variables whose attributes may be logically rank-ordered are ordinal measures. n 2022/1/3 Ranking operation: first, second, third…. . Example: Hardness of minerals § can determine the hardness of any two minerals by scratching one against the other. 2015 Research Methods (HK) 40

Ordinal Measurement with examples (con’t) Example: n the attributes of one’s leisure activities are: playing tennis, swimming, watching television, reading comics… n They can be ranked as follows: n 1. playing tennis (3) n 2. swimming (4) n 3. watching television (1) n 4. reading comics (2) 2022/1/3 2015 Research Methods (HK) 41

Ordinal Measurement with example (con’t) Cannot make any statements about the intervals between the numbers. The person may enjoy playing tennis nearly as much as swimming but find watching television is far more enjoyable than reading comics. What it implies is that the exact amount of difference is not known, it only indicates more or less, but not how much more or how much less. It is useful, when the researcher wishes to make judgment about one thing compared to another. The disadvantages are that it cannot perform most mathematical operations in analyzing data: not + - x ÷, can only do orderranking: 1<2, 2<3, and therefore 1<3. 2022/1/3 2015 Research Methods (HK) 42

Interval Measurement with example Has the characteristics of nominal and ordinal levels. Equal distance or intervals between numbers represent equal distances in the variable being measured. Example: Temperature scale (F) Difference between 20 F and 30 F is the same as the difference between 90 F and 100 F. n It can be concluded that 100 F is hotter than 90 F and also how much hotter it is. n 2022/1/3 2015 Research Methods (HK) 43

Interval Measurement with example (con’t) There is the establishment of a standard measurement unit: temperature (degree); time (hour, minute, second); length (feet/metre); income (dollars)…… When numbers represent a standard measurement unit, it can perform basic mathematical operations, that is + - , x ÷ . Interval measures do not have a true or absolute zero but an arbitrary one. Zero point on the scale does not signify the absence of the property being measured. 2022/1/3 2015 Research Methods (HK) 44

Interval Measurement with example (con’t) Example: Zero degree does not mean that there is no temperature. It is an arbitrary point on the scale. Its arbitrariness can be shown by comparison with another interval scale designed to measure the property of temperature: 0 F=18 C or 0 C=32 F Zero value does not mean one has found the least possible amount of the variable. One can get negative values at the interval level. 2022/1/3 2015 Research Methods (HK) 45

Ratio Measurement with example The highest and most precise level of measurement. Includes the features of the other levels plus an absolute zero point. For example, geographical distance, weight, time and height. Can distinguish among persons or objects by + - x ÷. 2022/1/3 2015 Research Methods (HK) 46

Ratio Measurement with example (con’t) Example: Income One earns $20, 000 and the other earns $40, 000 Can divide one into the other to signify that one is twice as much as the other. Zero point signifies none of the property being measures---age in years, number of siblings, year of schooling. n In operationalizing variables, this measurement measures at the highest possible level. n 2022/1/3 2015 Research Methods (HK) 47

Sources of Error Measures do not necessarily do a good job in reflecting what we intended to measure. Assessment of Reliability and Validity 2022/1/3 2015 Research Methods (HK) 48

Assessment of Reliability Definition “…reliability is a matter of whether a particular techniques, applied repeatedly to the same object, would yield the same result each time. ” Babbie E. 91 n 2022/1/3 Reliability is concerned with questions of stability and consistency. 2015 Research Methods (HK) 49

Assessment of Reliability (con’t) The kind of situations when reliability is in doubt: n a. ask the types of question people easily forget their answer. b. ask people the type of question that are irrelevant to them. n c. ask complicated questions that a person who had a clear opinion in the matter might arrive at a different interpretation of the question when asked a second time. n 2022/1/3 2015 Research Methods (HK) 50

Assessment of Validity Definition “Validity refers to the extend to which a specific measurement provides data that relate to commonly accepted meanings of a particular concept. ” Babbie E. 91 n The extent of matching, or goodness of fit between an operational definition it is purposed to measure. n 2022/1/3 The ability of a measuring instrument to accurately measure what it claims to measure. 2015 Research Methods (HK) 51

Assessment of Validity (con’t) To assess validity, it is necessary to ask the following questions: n a. Does this operational definition truly reflect what the concepts means? n b. Are you measuring what you intend to measure with this operational definition? If both the answers are “yes”, then you have a valid measure. 2022/1/3 2015 Research Methods (HK) 52

Summary of Section 2 You should have learned: . . most of the basic elements such as unit of analysis, time dimension of research. And be able to use in your research. . . how to construct a researchable research question. It is of great help to the planning of your research. . . how to state the hypotheses. They are to help you to make clear what variables are to be measure. . . what concepts that you are going to measure and how to measure them. Conceptualization is an exercise you must go through. . . the usage of the 4 levels of measurement. They are needed for the purposes of measuring as well as analyzing! 2022/1/3 2015 Research Methods (HK) 53

Readings Babbie, E. (2013) The Practice of Social Research. Chapters 4 & 5. Neuman W. L. (20130) Social Research Methods Chapter 6 and 7 (just the QN sections) 2022/1/3 2015 Research Methods (HK) 54