SOCI 3040 - Intro to Quant

Quantitative Research

Collects, analyzes and interprets numeric data. Counts, estimates, compares (in terms of size, intensity, perceptions, etc), and analyzes informations expressed through numbers.

• For example: Quantitative data is used a great deal in demography or the study of demographics (population, growth, change, nobility, etc)

Statistics Canada collects both…

Qualitative and quantitative information

Characteristics of Quantitative research

• Epistemological and ontological assumptions

• Collection of numerical data

• A deductive relationship between theory and research

• A preference for the natural science approach to research (positivism)

• An objectivist point of view

The 12 steps of Quantitative research

Theory → Hypothesis → Research design → Devise measures of concepts → Select research site (s) → Ethics review → select research participants → Administer research instruments/collect data → Process data → Analyze data → Findings/conclusions → Write up findings and conclusions.

What is true of the ethics process in quantitative research?

It is not optional. It may take as little as a few months but is more commonly known to take a long time— up to and over a year.

What is true of quantitative research as compared to qualitative research?

Quantitative research is quicker as a rule and data is simpler analyze.

Operationalization

The process of converting concepts into indicators or into specific questions in a questionnaire or an interview.

Dependent VS Independent Variables

• Dependent variable is the outcome and always depends on the independent variable.

• Independent variable is what is manipulated to change the outcome/dependent variable,

• For example: Headache pain measured based on caffeine intake. The headache pain is the dependent variable and it measured/determined based on the amount of caffeine consumed which the independent variable that changes and alters the outcome of the dependent variable.

Control variables

A variable the researchers assume is influencing the dependent variable, and thus is controlled for when testing the relationship between the independent and dependent variables. It is an extraneous variable that is controlled for.

• For example: A study that is evaluating the education level based on income earned. The dependent variable is the income ultimately earned, the independent variable is the education level that influences how much money you will ultimately make, and a possible controlled variable may be gender in this scenario because no matter education level, women are still paid less than men. This control is influencing the outcome of income earned (the dependent variable) however, it is not changing like the independent variable.

Reliability

Concerned with the consistency of measures. At least three different meanings:

1. Stability over time

2. Internal reliability

3. Inter-observer consistency

1. Stability over time

• One meaning of reliability

Whether the results of a measure fluctuate as time progresses, assuming that what is being measured is not changing. Stability can be measured using the test-retest method. It is very difficult to measure stability quantitively over time because of the number of factors that may come into play over the passage of time. The goal is to be able to retest this experiment and receive the same/similar answers.

2. Internal Reliability (or internal consistency)

• One meaning of reliability

Whether multiple measures that are administered in one sitting are consistent. This can be measured using Cronbach’s alpha coefficient or the split-half method. These calculations are completed using statistical programs.

3. Inter-observer consistency

• One meaning of reliability

All observers should classify behaviour or attitudes in the same way.

• For example: If 2 observers are recording the amount of aggression children display in a particular playground, their estimates should agree. If a child punched another child and one researcher says this aggression level is a 2/10 and the other says it’s a 8/10, there’s a problem. Researchers will have to go back to the drawing board to specify the criteria in which they’re measuring aggression level so that everyone conducting this research is measuring the same way,

Measurement validity

A measure that is not reliable will not be valid. An inconsistency in the way the data was gathered makes the data more or less unusable. A measure may be invalid but still be reliable. Data gathered for a research project may be invalid because it doesn’t “fit” with that particular project. But because it was gathered properly, it may be useful to inform a small facet of the research, provide a topic of further research, or provide useable data for other similar research.

Includes:

• Face validity

• Concurrent validity

• Construct validity

• Convergent validity

Face validity

Established if, at first glance, the measure appears to be valid.

• For example: At a coffee shop, asking people about their TV show consumption does not have great face validity. It’s no the appropriate setting, nor does the setting relate to the data being collected. A better question to ask in this setting would be surround coffee consumption, or caffeine consumption. This would have better face validity.

Concurrent Validity

Established if the measure correlates with some criterion thought to be relevant to the concept. A lack of correlation brings some doubt onto the validity (correctness) of the original measure.

Construct Validty

Established if the concepts relate to each other in a way that is consistent with the researcher’s theory, Confirmed by seeing that the results match what would be predicted given the theory.

Convergent Validity

Established if a measure of a concept correlates with a second measure of the concept that uses a different measurement technique.

Goals of quantitative research

• Measurement

• Establishing causality (internal validity)

• Generalization of findings to those not studied

• Representative sample

• Probability sampling

• Replication

Measurement

• One goal of Quantitative research

Data re used to understand or quantify social phenomena, concepts, and their interrelations in general.

Establishing causality (internal validity)

• One goal of quantitative research

Researchers want to know what causes social phenomena (ex: prejudice, crime, class conflict, etc).

Generalization of findings to those not studied

• One goal of quantitative research

The goal is to produce law-like findings that apply to large numbers of people (external validity). This is of a particular concern to researchers using cross-sectional and longitudinal design. Experimental model research is concerned more with internal validity than external. Having a representative sample is esstential for generalization (ex: not just conducting research on white men, include diverse and representative diversity that exists in the society you’re researching within).

Representative sample

• One goal of quantitative research

Representative only of the population from which is was taken (ex: a sample from a town only really applies to that specific town).

Probability sampling

• One goal of quantitative research

The use of a random sample drawn from a given population. Random selection does not guarantee representativeness of a sample group, but it does improve the chances.

Replication

• One goal of quantitative research

Provides a check for biases and routine errors. If the findings are not the same as those of the original study, the comparison provides reason to re-evaluate the methods and findings of the original study. If the findings are the same, researchers have greater confidence in the original findings.

Quantitative research within social sciences more broadly

• Mixed methods: incorporating components o quantitative work with qualitative work, it may fill in some gaps and give more accurate results to mis them together (ex: having people take a survey and then having a small interview with them afterwards).

• Use of survey data in addition to in depth interviews.

• Quantitative research is often privileged as “more scientific” or “ more objective” but ultimately, quantitative research is still suspect-able to some of the issues that qualitative research is plagued by.

• Quantitative research when done well can be extremely useful, intelligible, and helpful in evidence-based decision making.

Inductive VS Deductive Reasoning

• Inductive = Theory building - often found in qualitative research where researchers start with interviewing/observing/content analyzing to g enervate theories for why and how certain social phenomena happen. Bottom up reasoning. If three different student in sociology 3040 are studying it as a focus, then inductive reasoning would say that everyone in 3040 is focusing in it,

• Deductive = Theory testing/proving - Often found in quantitative research. Top down reasoning Taking a generalization and then finding specific evidence to support this generalization. If A = B and B = C, then A must also equal C. Example: Everyone in 3040 is in sociology, Jill is in 3040, therefore Jill is studying sociology.