Competency L

Demonstrate understanding of quantitative and qualitative research methods, the ability to design a research project, and the ability to evaluate and synthesize research literature.

Statement of Competency

A LIS professional with this competency knows how to conceptualize a research project by formulating a research question, planning how the research will be executed, measuring, categorizing, and analyzing that data to determine if the data answers the research question and supports the researcher’s theory.

The importance of understanding the research process is two-fold. LIS professionals consume research and conduct research. All fields of science need research including library and information science. Therefore, we need to do research to add to our field’s body of knowledge. Even though LIS is not pure science, our research can still be beneficial to address and solve problems in our field. And in order to make accurate observations, we need to approach our inquiries with scientific methods to make sure our conclusions are real and not subjective. Knowing how to conduct research is advantageous not only in the work place but also hones the researcher’s critical thinking skills because reading other people’s research makes one more perceptive and analytical.

Designing a Research Project

The research process is composed of three stages: problem identification, data collection, and data analysis. The first element of any research is an inquisitive mind so the process begins with a question that begs for an answer. This phase is called conceptualization. But in scientific or social research, one must first determine if the question or problem is worth studying.

Literature Review

Before research is conducted, it is prudent to review scholarly or research literature. Reading what has already been studied will give the researcher an opportunity to review findings from other studies, see if there are discrepancies between studies, or discover gaps in knowledge that requires further investigation. The literature review can be instrumental in identifying and refining the question the researcher wants to work on. In writing a literature review, a researcher should first provide a general overview of the research already done on the topic. The contents of this essay should cite works that are relevant to the research question, be organized by theme, and point out the areas where the proposed research can fill in gaps in knowledge.

Research projects can incur great expense so it should have significance or practical purpose. This is one of the advantages of conducting a literature review. It would be pointless to repeat research that has already been conducted. However, it does not necessarily mean that one cannot build on another person’s research.

Problem Identification

“The most difficult part of starting a research project is often that of identifying the best question to ask, one that is meaningful, whose answer contributes to the discipline, and whose resulting research can be carried with the resources available” (Black, 2002, p. 25). The question can be set up to support or refute a theory. On the other hand, theory can guide the questions the researcher asks and help understand the results from the study (SDCC, 2007, Paragraph 2). Research questions can either aim to replicate previous research, create comparative studies or start totally new research. Questions can be used to explore relationships between variables, describe the relationship between variables based on cause and effect, or explain why something happens between variables. Good research questions have certain qualities. They should be specific and can be evaluated with empirical, scientific methods. Avoiding vague research questions is key to providing direction for the research. More specific questions with a limited subset of variables have a better chance of being answered. Black (2002) wrote that a good problem statement should express a relationship between variables, be stated in a question form that is clear and unambiguous and imply the possibility of empirical testing (p. 26-27).

Data Collection

Once the topic of the research is established, the next part is operational stage. At this stage, decisions are made regarding data collection methods, sampling the population, and establishing parameters of the data collection such as budget, time constraints, and ethical guidelines.

Sampling is the primary method for selecting large representative sample for social research (Babbie, 2013, p. 186). A sample population represents the population that the researcher wants to study. It is almost always impossible to collect data from an entire population but getting a representative sample will allow for generalizations about that population. Major types of sampling include probability sampling, which is based on probability theory. Probability theory is the theory of analyzing and making statements concerning the probability of the occurrence of uncertain events (Dictionary.com, n.d., Paragraph 1). With probability sampling, a random representative sample is selected from a specified population, e.g. out of 100 persons, the researcher selects every 5th person to be a respondent. Researchers often prefer this method because they are considered more accurate and rigorous. Other advantages of probability sampling is that it avoids researchers’ biases in element selection and permits estimates of sampling error. Types of probability sampling include simple random sampling, equal probability of selection (EPSEM), probability proportionate to size sampling (PPS) and systematic sampling (Babbie, 2013, p. 190).

Non-probability sampling does not involve random sampling. It may or may not represent the study population. Yet there may be circumstances when it is not possible or practical to do random sampling and researchers have to rely on available subjects (Non-probability sampling, 2006, Paragraph 1). For example, interviewers may stop people on the street to answer questions. Non-probability sampling are more suited to qualitative research projects. Types of non-probability sampling types include purposive (judgmental), quota, and snowball sampling (Babbie, 2013, p. 186).

Conducting a survey is the most common way of collecting data. The survey can be in questionnaire form and this can be administered in paper or online form. Open-ended questions allow respondents to provide their own answers. This type of question are more appropriate for in-depth interviews. Close-ended questions limit the respondents to select an answer from a set provided by the researcher. These are typical in surveys because the responses are easier to process because they are more uniform, such as having just yes and no questions or an ordinal ranking of one to five. The most important thing to remember is to create questions that can be easily understood by respondents.

The experiment is another method of observation used by researchers. Experiments are usually conducted a controlled environment. An experiment has two stages: taking action and observing the consequences of that action. Experiments are appropriate for hypothesis testing, explaining cause and effect and small-group interaction studies. The classical experiment is the most common type of experiment and has three major components: 1) the independent and dependent variables, 2) pretesting and post-testing, and 3) experimental and control groups (Babbie, 2011, p. 230). Unlike the sample group which is representative of a larger population, it is more important that the people in the experimental and control groups are similar to each other. For example, the experimental and control groups are both composed of single, white females below the age of 30. Experiments can also be conducted as double-blind experiments, where neither the subjects nor researchers know which is the experimental or control group.

The major advantage of controlled experiments is the isolation of the variable’s impact over a period of time. Therefore subjects’ reactions can be attributed to the stimuli administered by researchers. Because of their methodical structure, experiments are also easier to replicate, even multiple times. Conversely, experiments are not a true reflection of real life. What occurs in a laboratory setting does not necessarily happen in natural social settings (Babbie, 2011, p. 250).

Evaluation research is a common form of research in social science. Evaluation research is not a research method per se but an application of research methods. “Evaluation can be a threatening activity” (RNKB, 2006, Paragraph 2). But evaluation is part of many companies and organizations, including libraries. Evaluations help these groups to determine if some social intervention is necessary or to rate the success of a particular activity or service. Although evaluation research is applicable to limitless topics, it is most appropriate for needs assessment studies, cost benefit studies, monitoring studies, and outcome assessment. All these are necessary in library operations and essential in securing funding. Evaluation researchers often conduct evaluations with experimental designs and quasi-experimental designs. Examples of quasi-experimental designs include time-series studies and the use of nonequivalent control groups.  Qualitative methods of data collection can be used but both quantitative and qualitative analyses can be used in evaluation research.

Key components to include when conducting an evaluation study are the specificity of the outcomes, interventions, population, units of analysis (or measurement), and there should be a definition of what success and failure look like. In the library setting, evaluation research can be done on a small scale, like how I aggregated the results of the quarterly library exit surveys for my workplace. All I needed to complete the data analysis was a tally sheet and Excel. For large-scale evaluations, there are evaluation services like LibQual®, a service provided by the Association of Research Libraries (LibQUAL, 2015, Paragraph 1).

Data Analysis

The final part of the research process is the measurement stage. In this stage, criteria are used to evaluate the quality of measurement based on precision and accuracy, reliability, and validity.  The results of data collection can be categorized as either quantitative or qualitative data. Babbie (2013) wrote, “The distinction between quantitative and qualitative data in social research is essentially the distinction between numerical and non-numerical data” (p. 23). There is no ironclad rule that a researcher has to choose between the two approaches. Many researchers use both although it is good to note that these two approaches need different skills and procedures.

Quantitative data is numerical, measurements are expressed in numbers such as age, income, household size, etc. Numerical data is easy to aggregate, compare and summarize. However, there is a potential loss in richness of meaning. The numerical descriptions cannot really capture the depth of language provide by human subjects. Quantitative research is useful for showing causality or how things became the way they are.  Another concern is generalizability (or external validity) which means that the result of the study can be applied beyond the sample population to a larger one. Finally, quantitative research is concerned with reliability (or internal validity) which means that a researcher can replicate a piece of the research and count on the reliability of the findings. Reliability tests produce the same results on successive trials (Analyze This, 2008, Sections 4/19 – 5/19).

Quantitative data analysis is the use of statistical techniques to describe and analyze variations in quantitative measures. In other words, the researcher will use numbers to discover and describe patterns in the data (Chambliss & Schutt, 2012, p. 154-155). Quantitative data analysis answers the questions of who, what, when, and where. It involves measuring or counting attributes (Analyze This, 2008, Section 1/19). Kruger (2003) states that “quantitative methods allows for summarization of vast sources of information and facilitate comparisons across categories and over time” (p. 18-19).

Qualitative data is non-numerical such as the description of a teacher’s strengths and weaknesses. Bawden (1990) defines qualitative data as “studying the behavior of individuals in all the complexity of their real-life situations” (p. 27). Qualitative data is a way of recording people’s attitudes, feelings, and behaviors in greater depth (Analyze This, 2008, Section 1/11). Qualitative data is richer in meaning but the purely verbal descriptions can also be a disadvantage because responses could be ambiguous or puzzling.

In qualitative data analysis, the focus is on the “text” extracted from transcripts of interviews or notes from observation sessions. “Text” can also refer to images that the researcher examines (Schutt, 2014, p. 321). Qualitative data analysis looks beyond numerical evidence and can be used to discover categories, themes, concepts and ideas within the data. Sources of qualitative data analysis include questionnaires, surveys, interviews, focus groups, observation and discourse analysis. Unlike quantitative data analysis, qualitative data analysis has the following features: it is often based on grounded theory practices; answers the ‘why?’ questions, and; pays greater attention to individual cases (Analyze This, 2008, Section 1/11).

Today, we have the capacity to send surveys all over the world, use text mining and data analytics software. As technology advances, LIS professionals are given more opportunities to leverage data to develop better programs, improve services, and project emerging trends. It is our responsibility to make sure that we keep our research and analysis skills updated to take advantage of this wealth of data.

Justification of Evidence

  1. Quantitative Data Analysis Quiz (LIBR 285 Research Methods)

This is a quiz on quantitative data analysis from my research methods class. I included this artifact because it is proof that I understand quantitative data analysis. I was able to identify the four types of variables: nominal (names or labels), ordinal (rank order), interval (scaled), and ratio.  I was also able to define “standard deviation” which is a measure of dispersion around the mean and the standard error of a sampling distribution. I also measured the central tendency using a given data set. Based on the information given, I determined that the best measure for central tendency to use was median because the data was a set of ages from 1 to 88 years and the median is not affected by how far away from the middle values are. The last question asked that I define a bivariate table and enumerate the steps on how to construct one. The second part of the question involved creating and interpreting a bivariate table based on the data provided in the word problem.

  1. Qualitative Research Presentation (LIBR 285 Research Methods)

This presentation was also for my research methods class. I had to critique an article by Jennifer Waugh on formality of chat reference. The purpose of the qualitative study was to determine whether the formality of language used by librarians affect 17- to 25-year old university students’ perceptions of synchronous virtual reference interactions or more commonly known as “chat reference.” Waugh examined two transcripts, one having traditional, formal language and a second with more informal language. The sample group was composed of 5 university students within the specified age group. I identified the four questions used by Waugh to test her hypothesis. I also examined her literature review where she evaluated the studies done by Dewdney and Ross on interpersonal communication and Mons study on users’ response to friendly and polite librarians. She further explored the studies of politeness theory by Brown and Levinson and her investigation suggest that formality is a critical factor in virtual reference interactions. This is an example of discourse analysis (which is an examination of oral, written or sign language).

  1. Research Proposal (LIBR 285 Research Methods)

I’ve included my final project for my research methods class which is a research proposal on a topic of my choosing. The subject of my proposal is the culture of anonymity practiced by librarians and its effect on reference user satisfaction. The premise of the proposal is that other service professional are easily identifiable by name tags (e.g. nurses) and clients prefer establishing a standing relationship with these professionals. However, this is not the case in libraries. In fact, library cannot distinguish between the different library workers. So this became the basis of my research problem. Before proceeding with the proposal, I did a literature review to see if this issue has been previously and sufficiently addressed. I discovered different reasons for this practice, its effects, and arguments for self-disclosure. I also wrote the methodology for this proposal which specified the target sample, method of data collection and analysis, and research schedule. I also included researcher qualifications, the significance of the research, a summary, and references. The appendices include a request to conduct research, field notes guide, field observation form, proof of consent, and a survey questionnaire. The proposal can be used to request permission to begin a research project or secure funding like a grant. But it is also a good exercise to see if the research problem is an issue that is worth pursuing and if the researcher can contribute new knowledge to the topic.

Evidence

Quantitative Analysis Quiz

Qualitative Research Presentation

Research Proposal

References

Analyze This. (2008). Qualitative data analysis. Retrieved from http://archive.learnhigher.ac.uk/analysethis/main/qualitative.html

Analyze This. (2008). Quantitative data analysis. Retrieved from http://archive.learnhigher.ac.uk/analysethis/main/quantitative.html

Babbie, E. (2011). The practice of social research. 13th ed.

Babbie, E. (2013). The basics of social research. 6th ed. Boston, MA: Cengage Learning.

Bawden, D. (1990). User-oriented evaluation of information systems and services. Surrey, UK: Gower.

Black, T. (2002). Questions and hypotheses. In Understanding Social Science Research. London, UK: Sage.

Chambliss, D.F. & Schutt, R.K. (2012). Making sense of the social world: Methods of investigation. 4th ed. London, UK: Sage.

Kruger, D.J. (2003). Integrating quantitative and qualitative methods in community research. The Community Psychologist, 36, 18-19.

LibQUAL+®. (2015). What is LibQUAL+®? Retrieved from https://www.libqual.org/about/about_lq/general_info

Probability-theory. (n.d.). Dictionary.com Unabridged. Retrieved from http://dictionary.reference.com/browse/probability-theory

RMKB. (2006). Evaluation research. Research Methods Knowledge Base. Retrieved from http://www.socialresearchmethods.net/kb/evaluation.php

RMKB. (2006). Nonprobability sampling. Research Methods Knowledge Base. Retrieved from http://www.socialresearchmethods.net/kb/evaluation.phphttp://www.socialresearchmethods.net/kb/sampnon.php

RMKB. (2006). Probability sampling. Research Methods Knowledge Base. Retrieved from http://www.socialresearchmethods.net/kb/sampprob.php

Schutt, R.K. (2014). Investigating the social world. 8th ed. London, UK: Sage.

The Sociology Department of Colorado College. (2007). Good research questions. Retrieved from https://faculty1.coloradocollege.edu/~mduncombe/Statistics/quanthome/Question/goodq.htm