The Nature of Quantitative research

The Nature of Quantitative Research Lecture #4 Chapter 3 February 4, 2013 • Quantitative research entails the collection of numerical data, a deductive (theory- conclusion) relationship between theory and research, a preference for the natural science approach to research (positivism in particular), and an objectivist conception of social reality. Key terms • Concepts: ideas or mental representations of things - E.g.: crime, gender, alienation, gender, life satisfaction - Concepts may be independent or dependent variables. For example, the concept ‘social mobility can be used as a possible explanation of certain attitudes or as something to be explained • Coding: transforming a measure into numbers - Can be done in both qualitative and quantitative data - E.g.: in measuring life satisfaction, respondents who say they are “very satisfied” may be given a code of “1”, which is then recorded in a file. - Every data set has a coding manual Why measure concepts? - Measurement allows a delineation of small or fine differences between people or issues we are interested in - Measurements provide a way to identify and gauge those differences with consistency - It allows us to estimate what the relationship is between concepts and the strength of that relationship. What is the correlation? Is it positive or negative? Two types of definitions for a concept: • Nominal: describes the concept in words, much like a dictionary definition, E.g.: political party identification refers to a party to which people most closely associate themselves • Operational: describes how the concept is to be measured. E.g.: political party identification may be measure by asking people, “do you normally think of yourself as a supporter of the Conservatives, Liberals, NDP, or Bloc Québécois? • Indicators tell us (indicate) that there may be a link and tell us how strong the link may be • Usually, one indicator for each concept is adequate • It is advantageous to use more than one indicator of a concept (often done in survey research) because it: - Reduces the likelihood of misclassifying some people because the language of a question is misunderstood - Ensures the definition of the underlying concept is understood correctly - Gets access to a wider range of issues related to the concept, allows the researcher to make finer distinctions, and - Allows factor analysis (the process of sorting out the degree to which a particular concept is being measure and by which indicator) - E.g.: employee moral; how much does one love his or her job? Can look at attendance, can be an indication of how much one likes their job Dimensions of Concepts - in developing a measure for a concept, its different aspects should be considered • 1. Reliability: ultimately refers to stability of a measure over time, i.e.: whether one would get the same result after using a measure at two different times (assuming no change in what is being measured). - Stability can be measured using the test- retest method. - It is extremely difficult to quantify stability over time because of the number of factors that may come into play over the passage of time, including the fact the respondents took a prior or pre-test and that may influence subsequent tests; more stable if done with different groups in different times. • 2. Internal reliability/ internal consistency: are multiple measures administered in one sitting consistent with each other, e.g., multiple measures of “life satisfaction” on a survey. - This can be measured using Cronbach’s alpha coefficient or the split- half method. These calculations are done using statistical programs. - Acorrelation of .8 or higher on a scale of ‘0-1’is generally accepted as a minimum of internal reliability, although results with lower figures may still be used by some researchers. • 3. Inter- observer consistency: all observers should classify behaviors or attitudes the same way - e.g.: if two observers are recording the amount of aggression children display on a playground, their estimates should agree • Measurement validity: concerned with whether one is measuring what one wants to measure - There are various kinds of measurement validity: 1. Face validity: is established if, at first glance, the measure appears to be valid. Ask for the opinion of someone with expertise in the area of study 2. Concurrent validity: is established if the measure correlates with some criterion thought to be relevant to the concept.Alack of correlation beings some doubt onto the validity (correctness) of the original measures - E.g., if IQ scores correlate as expected with grades in university, concurrent validity is established 3. Construct validity: is established if the concepts relate to each other in a way that is consistent