2 Survey Research Process: Measurement 3 Types of Variables 4 Types of variables Explanatory vs. Extraneous Those variables which are selected for analysis versus those which are not due to their perceived relevance to the research question 5 Types of variables Explanatory variables Dependent or criterion variables Independent or predictor variables Control variables 6 Types of variables Latent vs. Manifest Variables Latent variables are not directly observable or measured; they must be observed or measured indirectly, hence inferred e.g. Intelligence, Degree of decentralization, Socio-economic status, Conservatism Manifest variables, also known as observed or indicator variables, can be directly observed or measured e.g. Score on tests, Ratio of supervisors to workers, Income level, Employment rate, Literacy rate, Value orientation Levels of Measurements 7 Measurement Measurement is the process of assigning numbers or labels to the attributes of objects, persons, states, or events in accordance with specific rules 8 Levels of Measurements 9 http://www.socialresearchmethods.net/kb/measlevl.php Levels of Measurements Knowing the level of measurement helps in deciding how to interpret the data from that variable Knowing the level of measurement helps decide the statistical analysis appropriate on the values that are assigned 10 11 Levels of measurements Nominal Variable or Nominally Scaled Variable Has all the features of a variable but no other mathematical property Classifies objects into distinct groups e.g. Sex (Male, Female), Religion (Catholic, J ews, and Other), Ethnic origin (North American, European, African, South Asian, Other) 12 Levels of measurement Ordinal Variables The categories that partition a variable differ from one another in quantity of an attribute and not just quality These categories have an order and can be arranged from the smallest quantity of the attribute to the largest. But, the distance between categories has no meaning Contain all properties of the nominal variable with additional property of ordering e.g. Social Class, Educational attainment 0: Less than grade 8; 1: Grades 8-12; 3: Some college; 4: College degree; 5: Post college 13 Levels of measurement Interval Variables Categories correspond to numbers and distance between the categories has meaning that can be interpreted Have all properties of ordinal variables (classification and ordering) with additional property of interval But, ratios don't make any sense e.g. Temperature in Fahrenheit 14 Levels of measurement Ratio variables Has all the properties of the previous measurement and in addition there is always an absolute zero that is meaningful We can construct a meaningful fraction (or ratio) with a ratio variable e.g. Age, Weight, Income 15 Hierarchy in levels of measurements 16 Levels of measurement Discrete vs. Continuous Variables A discrete variable is a scale in which the categories are disconnected values, the interval between adjacent categories cannot be subdivided. In other words, a discrete variable can assume a finite number of values e.g., Household size, Class size, Number of houses, Number of heads in ten tosses of coin 17 Levels of measurement In contrast to discrete measures, continuous variables can, at least in theory, assume all possible numerical values in a given interval. Many attitudinal variables which are ordered discrete variables, are treated as continuous variables e.g. Income, Age, Height, Weight 18 Process of Measurement Developing Indicators 19 Survey Data Case by variable Matrix RESP_ID SEX AGE EDUCATION TIMBER RECREATION 001 Male 36 Years College deg. Agree Disagree 002 Female 40 Years High School Agree Disagree 003 Female 75 Years <Grade 8 Disagree St. Agree 004 Male 25 Years Some Univ. Cant say Disagree 005 Female 32 Years High School St. Agree Cant say 006 Male 58 Years Grade 13 Cant say Agree 007 Female 23 Years Diploma St. Disagree Agree Variables Cases Identify the concept of interest Develop a construct A constitutive definition An operational definition A measurement scale Evaluate the reliability and validity of the scale Utilize the scale Research findings The Measurement Process which is used to create which enables a researcher to develop which enables a researcher to create that requires the researcher to If the evaluation is satisfactory, the researcher which leads to which is used to http://malroy.econ.ox.ac.uk/fisher Identify Concept / Develop Construct Identify the concept of interest Develop a construct A constitutive definition An operational definition A measurement scale Evaluate the reliability and validity of the scale Utilize the scale Research findings Identify Concept / Develop Construct A concept is expressed in every-day terminology. This requires the researcher to generalize/categorize. is an abstraction formed by generalization from similar phenomenon or similar attributes A construct is a theoretical abstraction that cant really be observed is a concept that is systematically defined to be used in scientific theory 23 24 Research Question Why do Rural Customers buy Lifebuoy Soap over and over again? Identify Concept / Develop Construct Concept: Repeat Purchase Construct: Brand Loyalty 25 26 Clarifying Concepts/ Constructs Abstract summaries of a whole set of behaviors, attitudes and characteristics Are terms created for the purpose of communication and efficiency Lack fixed meaning The challenge is to develop concepts/constructs that others will clearly understand It is up to the researcher to first define the concept with some justifications, and then look for constructs e.g., Deprivation, Social capital, Community, Brand Loyalty, Religiosity 27 Clarifying Concepts/ Constructs Obtain a range of definitions In the absence of formal definitions look at the meaning the term has been used by various researchers explicit vs. implicit definition Look at common elements of definitions and based on these, develop a definition 28 Clarifying Concepts/ Constructs Decide on a definition Choose an existing definition or Create your own In either case, justification is needed Defining Concepts/ Constructs 29 Constitutive Definition A theoretical or conceptual definition that defines the concept in terms of other concepts and constructs; like a dictionary definition Operational Definition Defines which observable characteristics will be measured and the process for assigning a value to the concept Define the Concept Conceptually and Operationally Constitutive Definition Increased propensity to purchase a brand due to previous experience with that brand Operational Definition Rating of purchase probability, depending upon prior purchase Example 31 32 Socio-demographic conditions influence ones shopping behaviour 33 Conceptual propositions Older people have conservative shopping behavior Educated people show progressive shopping behavior Similarly Parents level of education Gender Place of residence 34 Operational Definitions Operational Definition of Key Concepts & Identification of Indicators Younger generation: People in the age group of 15-34 Years Older generation: People in the age group of 65 years and over Higher level of education: University/College degree or above Urban areas: Places above a specified population density 35 Operational Definitions Rural areas: It has a population density of less than 400 sq. km. At least 75 per cent of male working population is engaged in agriculture related activities, and There is, in the location, no municipality or municipal board Shopping behavior: Conservative orientation Progressive orientation 36 Deprivation Nominal Definition Physical Economic Social Political Dimensions Psychic Social Isolation Ab. of soc. valued roles Lack of social skills Sub- Dimensions Specify what constitutes Social Isolation Specify what Soc. Valued roles are Specify what Social Skills are Operational Definition Measures: e.g. No. of friends Contact with family Contact with neighbors No. of org. involvement Measures: e.g. Occupation Gender Measures: e.g. Introversion/ Extroversion Assertiveness Indicator Clarifying Concepts Descending the Ladder of Abstraction (Source: de Vaus, 1991)