Key Concepts Blog: Research Traditions Readings

Williamson (2000) Chapter 2

The Two Major Traditions of Research
Positivist Research

Philosophy – all scientific knowledge should be based on objectively observed experience.

Usually linked with deductive reasoning, quantitative methods (may also include qualitative especially by post-positivists).  It seeks to link cause and effect and generalizations eventually made.  Researcher tries to refute his/her hypothesis.  If data is consistent with hypothesis, the hypothesis is corroborated (i.e. temporarily supported, but not proved).

Post positivists use more natural setting for research, seek more insider (emic) views) and incorporate more qualitative methods.

Interpretivist Research

Philosophy – social world is interpreted or constructed (constructivism).  It is concerned with meaning.

It is mainly associated with deductive reasoning and qualitative methods.  It uses naturalistic inquiry.  They plan their research, but less linear.  The researcher may make adjustments in light of new perspectives during the research.  Emphasis is not given on generalization.  However, it needs to be rigorous and validity and reliability are insured by comparing with literature and triangulation.

Pyrczak (2008) – Appendix A
Quantitative and Qualitative Research

Steps in Quantitative Research

·       Start with hypothesis/purposes derived from previous research or theory.

·       Select random sample representative to population (relatively large sample of participants).

·       Make observations with instruments that can be scored objectively.

·       Describe results using statistics and make inferences to population.

Quantitative researcher has limited interaction with participants.

Steps in Qualitative Research

·       Start as general research question, not hypothesis.

·       Select purposive sample that the researcher believes to be appropriate, not random and not necessarily representative of larger population.

·       Relatively small sample like one exemplary case or classroom.

·       Observe with relatively unstructured instruments like semi-structured interviews or unstructured observations.

·       Observe intensively for extended period of time to get in-depth insight.

·       Results presented in words with emphasis on the sample, without generalizations to larger population.

It is personal, interactive and characterized by researcher’s awareness of their own orientations, biases, experiences.

Connaway and Powell (2007) Chapter 2

Developing the Research Study
Identification of the Problem

·       See and state the problem clearly/precisely.

·       It may be from previous related research in the field and gaps in knowledge.

·       Steps

o   Write clear statement of the problem in complete grammatical sentences.

o   Identify and write sub problems, that when combined, are equal to the whole.

Hypothesis

Hypothesis is “tentative propositions set forth to assist in guiding the investigation of a problem or to provide possible explanations for the observations made.”

An ideal hypothesis should be:

·       Generalizable, universal, i.e. hold up in more than one situation.  It may also be for a specific situation.

·       Compatible with existing knowledge/not isolated from larger body of knowledge.

·       Testable.

Eldredge (2004)

Inventory of Research Methods for Librarianship and Informatics
Cohort Design

Tracks overtime a defined population that share common characteristics as it encounters a phenomenon

The exposure may be intended or unintended.

Eg. Students’ information literacy before and after they are exposed to library or information instruction

Three types of cohort studies

Retrospective – when study begins after exposure.

Prospective – when study begins before exposure.

Longitudinal – when multiple measurements are taken at regular intervals.

Case Study

Investigating a contemporary phenomenon in its real-life context … utilizing multiple sources of evidence

Describes and analyzes the author’s experiences with a process, group, innovation, technology, project, population, program or organization.

Patten (2000) Topics 1-10

Topic 1 

Empirical Research - is based on observations, but differs from everyday observations because it is carefully planned. 

Planning the Research – empirical scientist try to avoid misleading results and poor interpretations by carefully planning why (the research need) and whom (sample or whole population) they observe, and how and when (research instruments) to observe.

Topic 2

Experimental research is when the researcher gives treatments and observes if they cause any changes.  Usually, in an experiment, the group to be studied is divided into two randomly and treatment is given to one group (experimental group) and no treatment is given to the other group (control group).  Sometimes, groups may not be assigned at random, but it may still be experimental research.  However, the control group should have the same/comparable demographics, background etc. as the experimental group.

Non-experimental research observes subjects in order to describe them as they naturally exist without giving any treatment.  Some kinds of non-experimental research are casual comparative research, survey/poll, census, case study, field research/ethnographic research, longitudinal research, correlational research, and historical research. 

Topic 3

Casual-Comparative Study (ex post facto study) is non-experimental research which is done when experimental research cannot be done due to physical, ethical, legal and financial reasons.

Steps in casual comparative study

• ·       Observe and describe current condition
• ·       Look to the past and try to identify possible causes

Topic 4

Field research/ethnographic research is a kind of case study where we study a group while becoming a member of the group.

Correlational research repeatedly measures trait(s) of subjects over a period of time to trace developmental trends.

Topic 5

A variable is a trait or characteristic with two or more categories.  (Example – variable - gender, categories – male or female.)  All variables have mutually exclusive categories, i.e. each subject belongs to only one category.

Types of Variables 

• Naming/categorical variables (example - gender – name or female) versus quantitative variables – (example - GPA)  
• Dependent versus independent variables.

Topic 6

All experiments have at least one independent variable (stimulus, input) and one dependent variable (response, output).  But they may have more than one independent variables and/or more than one dependent variables.

The independent variable is physically manipulated to observe the extent to which it causes changes in the dependent variable.

Topic 7

Hypothesis is educated guess or formal theory.  It is prediction of outcome of the study.  It may be directional or non-directional.

Research Purpose and Research Question - Hypothesis may instead be expressed as research purpose or research question.

Topic 8

Conceptual definition is definition of concepts of terms as in dictionaries.

Operational definition indicates physical steps we take to observe the variables.  Operational definition should be adequate enough to permit replication of the research.  It should also be meaningful and relevant.

Topic 9

 Quantitative Research

• ·       Deductive research - starts with literature review and deriving hypothesis from literature.
• ·       Uses highly structured instruments and produce numbers.
• ·       Large sample.
• ·       Limited interaction with subjects.
• ·       Doesn’t make adjustments in the research.
• ·       Summarizes all responses and doesn’t report on individuals.
• ·       Generalizes the report to one or more population.

Qualitative Research

• ·       Inductive – hypothesis from preliminary study of the subjects.
• ·       Unstructured instruments and produce words.
• ·       Small sample.
• ·       Spend a lot of time interviewing and observing.
• ·       Makes adjustments as necessary during the research.
• ·       Tends to cite individuals.
• ·       Limits findings to the individuals/group studied.

Topic 10

 Favor qualitative research when:

• ·       the topic warrants qualitative research (example – emotional impact …)
• ·       little is known about a topic.
• ·       in secretive cultures.
• ·       time and funds are limited.

Favor quantitative research when:

• ·       the topic warrants quantitative research (example – economic conditions …)
• ·       subjects are not available for extended interview/observation.
• ·       numbers are required.

A blend of qualitative and quantitative methods may be used.

Sources

Connaway, L. S.  and Powell, R. (2007). Chapter 2. In Basic Research Methods or Librarians (5th ed.). California.

Eldredge, J. D. (2004). Inventory of Research Methods for Librarianship and Informatics. Journal of the Medical Library Association, 92(1), 83.

Patten. (2000). Topic 1 - 10. In Understanding Research Methods. Pyrezak Publishing.

Pyrezak, F. (2008). Appendix A. In Evalutaion Research in Academic Journal. Pyrezak Publishing.

Williamson, K. (2000). Chapter 2. In Research Methods fr Students and Professionals: Information Management and Systems.