Week 7--Key Concepts in Research

Key Concepts in Research—Week 7

Data Analysis Techniques

Connaway & Powell—Chapter 9

Dispersion or variability—in statistics, this refers to how widely the cases vary within a group. The four descriptive statistics commonly used for this are: range of scores, mean deviation, standard deviation, and variance.

What are the basic steps of statistical analysis? 1) Establishment of categories 2) coding the data 3) analyzing the data using either descriptive or inferential statistics.

Wildemuth—Chapter 29

Content analysis—is a research method often used in journalism, mass communication and library and information science. The method is a quantitative analysis of words that are systematically pulled from texts, graphics and video, to objectively examine the contents.

Intercoder agreement—when more than one coder is coding the data it is important that they are objectively coding in the same manner. To test the agreement, first a measure of agreement is conducted, second choose the minimum amount of reliability that is acceptable in the measure of agreement, third assess informally during the training of coders, and formally in a pilot test and again during the actual coding of the research. Last, the results from your intercoder agreement must be reported in your research write-up.

Wildemuth—Chapter 30

What are the steps of qualitative content analysis? 1)Prepare the data 2) Define the unit of analysis 3) Develop categories and a coding scheme 4) Test the coding scheme on a text sample 5) Code all of the text 6) Assess the consistency of the coding 7) Draw conclusions from the coded data 8) Report the methods you used and the findings.

What are the three types of qualitative content analysis discussed by Hsieh and Shannon (2005)? The three types are conventional qualitative content analysis, directed content analysis, and summative content analysis.

Wildemuth—Chapter 31

Discourse analysis—is used to analyze text and speech to uncover meaning in the selection of words, expressions, hesitations, etc. It makes the assumption that text and speech are what builds the speakers’ social world.

Hermeneutics—is a method of analysis in which a text is broken down into parts, and looked at separately, and then analyzed for how the parts affect the interpretation of the whole.

Wildemuth—Chapter 32

Analytic induction—is a method used when analyzing qualitative data. Using the data, the researcher develops and refines a theory or hypothesis.

Moderatum generalization—analysis is a modified qualitative analysis that generalizes analytic induction in a more even-handed manner than purely qualitative analysis.

Wildemuth—Chapter 33

Descriptive statistics—summarizes the results of your data. Various statistics may be used, including frequency distribution which can be displayed in contingency tables and discussed using cross-tabulation and chi-square analysis.

What are two types of statistics frequently used in descriptive statistics? Measures of central tendency include the mean, the median, and the mode and are concerned with one value that best summarizes the data set. Measures of dispersion describe your data set as a whole. When using the mode or the median, the range or interquartile range is used. If the mean is used, report the standard deviation.

Wildemuth—Chapter 34

Frequency distribution—is an organized table displaying the number of cases in a specific category of a specific variable.

Chi-square statistic—a test used to determine if the frequency distribution is likely to occur by chance. It measures what one observes and what one expects to observe in the general population.

Wildemuth—Chapter 35

Analyzing sequence of events—is often used to determine behavior in web browsing and searching in library and information science. The researcher is interested in the sequence that a subject uses during a certain behavior such as dancing, channel surfing, web browsing, etc.

How is sequential data analyzed? By using state transitions where observation is made from one event to the next or by using optimal matching algorithms which attempts to compare to complete sequences looking for similarity or dissimilarity.

Wildemuth—Chapter 36

Correlation—a method of statistical analysis used to examine the relationship between two variables. A scatter diagram can be used to graphically display correlation.

Strength of relationship—the absolute value of the correlation statistic, it will range between 0 and 1. The higher the absolute value is indicates a stronger relationship between the two variables.

Wildemuth—Chapter 37

T test—calculates the probability that the difference between two means is attributable to chance.

Analysis of variance (ANOVA)-similar to the t test the ANOVA can be used to test more than two means.

References

Connaway, L. & Powell, R. (2010). Basic research methods for librarians. 5th ed. Santa Barbara, CA: Libraries Unlimited.

Wildemuth, B. M. (2009). Applications of Social Research Methods to Questions in Information and Library Science. Westport, CN: Libraries Unlimited.

Week 6--Data Collection Techniques

Key Concepts in Research—Week 6

Data Collection Techniques

Connaway & Powell—Chapter 5

Questionnaire—is often used to gather quantitative data. The questionnaire must be carefully planned to gather the information that will help the researcher answer the hypothesis. Advantages of using a questionnaire as a data collection instrument include: frank answers due to the fact that anonymity is possible because they are often via mail, email, or web-based forms, interviewer bias is eliminated, each question is asked the same without interviewer bias, participants can answer the questionnaire at a time that is convenient to them, easy to collect quantitative data, a large number of participants can be included, and the cost is less than many other techniques. Disadvantages of using a questionnaire include: personal contact between researcher and participant is lost, participants cannot ask the researcher questions for understanding, strongly opinionated people are more likely to answer surveys which can skew the results, less educated individuals may have difficulty participating causing bias towards more educated respondents, some participants may purposely sabotage the survey by giving extreme answers, nonresponse rates are high, and if distributed via internet it will only be available to people with access to the internet.

Observation—is considered both a method and a technique and is central to qualitative research, but is also used to gather data for quantitative research. The researcher closely observes participants in a scientific or systematic way. Advantages of observation research include: behavior can be recorded as it occurs, comparison can be made between what participants did and what they said they did, all behavior and actions can be recorded by the researcher eliminating the participants opinion of importance, anyone can be observed regardless of verbal skills, and observation can be accomplished without participants knowledge, however there are ethical implications that need to be considered. Disadvantages of observation research include: it is difficult to always be available to observe unstructured events, the length of the event may be too brief to observe, and it can be difficult to quantify observational data.

Wildemuth—Chapter 18

Transaction logs—computer records that record data as applications are being used. Records that can be extracted from logs include: IP address, date and time of page requests, which files were requested, web server logs, use of search engines, just to name a few. This data is then used to help researchers better understand search behaviors and interpersonal interactions in interactive features such as chat.

The process of transaction log analysis—there are three main steps to processing transaction logs for analysis. First, one must identify the data elements and sources appropriate to the study. Second, the data must be collected. Third, the data must be cleansed and prepared for analysis.

Wildemuth—Chapter 19

Think aloud protocols—also referred to as verbal reports and verbal protocols. The study is set up for the participant to think out loud while performing a task or solving a problem.

Concurrent and Retrospective Protocols—in concurrent protocols, the participant thinks aloud while performing a task or solving a problem. In retrospective protocols, the participant reports what he was thinking while he was performing a task or solving a problem after he is finished with the task.

Wildemuth—Chapter 20

Direct observation—allows participants to be observed or questioned during an event which helps the researcher get the full picture of timing, duration or frequency of behaviors. The participants may not be fully aware of the researcher due to the event occurring, allowing for realistic activities to naturally occur.

Participant and nonparticipant observation—in participant observation, the researcher has the dual role of being a participant in the study while also observing. In nonparticipant observation, the researcher is only an observer.

Wildemuth—Chapter 21

Advantage to participant observation—may lead to better understanding of the participants and social processes that occur which may lead to better theories.

Researcher role in participant observation—the researcher’s role may be passive or limited where the researcher stays in the background as much as possible. If the involvement is complete, then the researcher is a true member of the culture. This is referred to as full membership or active participant.

Wildemuth—Chapter 22

Research Diary—are varied in form and content. In research, the author may be asked to record data of events in a log, or may be asked to write narrative personal accounts. The information may be used in quantitative or qualitative research.

Types of diaries—three types of research diaries are identified in the Wildemuth text. Unstructured diaries are open-ended journals that give the participant little if any instruction regarding the content or form that is expected of the diary. Semistructured diaries fall along the middle of the continuum between structured and unstructured diaries and are probably the most frequently used. Structured diaries are very specific in the data and form that the researcher instructs the participant to record.

Wildemuth—Chapter 23

Unstructured Interviews—a technique of interviewing developed in the anthropology and sociology disciplines. It is also referred to as informal conversational interview, nonstandardized interview, ethnographic interview, or in-depth interview. The researchers who use this method generally have a constructivist point of view and are working to interpret the social reality of participants.

Aide-mémoire—an agenda used for unstructured interviews. It gives the interviewer broad guidelines on the topics that may be covered in the interview.

Wildemuth—Chapter 24

Double-barreled question—when two issues are covered in one question during an interview. This can make it difficult for the respondent to clearly answer the question and for the researcher to analyze the response.

Time-line interviewing—a technique such as the micromoment time-line interviewing, commonly associated with Dervin’s sense-making theory. A time-line of events is established and then questions are asked about each event in the time-line.

Wildemuth—Chapter 25

Focus group—the researcher gathers a group of individuals together to discuss the topic of research. Participants are able to compare their view with others in the group and discuss and challenge views, giving researchers a more nuanced perspective of the topic.

Control characteristics—the characteristics that you would like participants in your focus group to possess.

Wildemuth—Chapter 26

Survey instrument—how the questions of the survey will be formulated to generate responses. Careful drafting is imperative.

Methods to increase response rates—contact prospective respondents multiple times, collaborate with a relevant organization, get individual to sponsor the study, or offer an incentive.

Wildemuth—Chapter 27

Construct—is the object of measurement in a study. It is more complex than just a simple question. A construct must be defined and then operationalized.

Inventory—is developed through defining cognitive and affective variables. Many are identified through previous studies done by other individuals.

Wildemuth—Chapter 28

Likert Scale—an often used format which measures psychological constructs. A statement is rated on a 5-point scale rating the agreement with the statement between strongly agrees or strongly disagrees.

Semantic Differential Scales—a method of measuring connotative meaning of objects, events, people, etc. to the individual responding to the scale. The participant rates opposites along a continuum expressing their attitude towards the stimulus.

References

Connaway, L. & Powell, R. (2010). Basic research methods for librarians. 5th ed. Santa Barbara, CA: Libraries Unlimited.

Wildemuth, B. M. (2009). Applications of Social Research Methods to Questions in Information and Library Science. Westport, CN: Libraries Unlimited.

Article 2 Review-Citation Analysis

Diane Mick

LIS 7996: Research

October 6, 2011

Clark, S. & Chinburg, S. (2010). Research performance in undergraduates receiving face to face versus online library instruction: a citation analysis. Journal of Library Administration, 50 (6), 530-422 doi: 10.1080/01930826.2010.488599

As online college programs continue to increase in availability, it is important to improve the delivery methods and analyze the quality of education students receive. For Library and Information Science, it is time to integrate the knowledge of search skills into classes to increase the quality of articles and research cited by students. The question I would like to answer in my research is if the quality of education is improved by the embedded librarian program offered in some online classes. Unfortunately, this program is not available at all colleges and further research could persuade administrators to provide this service in all online academic programs.

In 2008, Rogers State University, Claremore, Oklahoma began offering the embedded librarian program into online courses if requested by instructors. Susan Clark and Susan Chinburg launched a citation analysis to test the effectiveness of the embedded librarian program.

The Clark and Chinburg study included three sections of an upper-level Management Information Systems (MIS) class. All three sections had the same instructor. Two sections were offered online, and one section was a face to face class that met twice per week. Each class covered the same material and used the same textbook. The research papers evaluated were from two semesters for a total of six sections. The results compared 327 citations.

The independent variable is the research papers written by all students in this class. The dependent variable is if the student was online or face to face. The face to face section was given one 50-minute class period of traditional information literacy instruction by the access services and distance learning librarian. The online sections were provided with a PowerPoint tutorial with voice over and screen shots by the same librarian. Links to key resources and the same information was given to both the face to face and online students. The librarian was then embedded into the online course with teaching assistant privileges in the course-management software. The embedded librarian read of the discussion board post and had an area designated for research questions which were answered with appropriate responses and suggestions throughout the semester. The research paper instructions for all sections were identical. The students were to write a strategic analysis paper, six to nine pages in length, on a technical topic in the information systems field. They were required to use five sources with no more than three sources from Internet-only articles.

To begin analyzing the sources, the citations were stripped from the papers by the professor and forwarded to the librarian. The citations were categorized into eight types of print and electronic sources. The citations were then tested against the null hypothesis that there would be no significant differences, statistically, between the sources used in face to face and online sections of the class. The class section results were very similar with no statistically significant differences. A chi-square test was used to test the results against the expected null hypothesis. The results were well within range with only a variation between the two datasets of 0.963.

The study was successful in suggesting that online instruction can be as effective as face to face instruction if an embedded librarian is used in the online class and research instruction is given to all sections. The study would have been more helpful if two more sections, face to face and online, without library instruction would have been tested along side those with library instruction.

I would suggest this article as a starting point for someone wanting to use citation analysis in a quantitative study. More studies need to be conducted to improve online classes and validate online education as an equally effective means of education.

Week 5--Methods/Ethics/Theory Readings

Key Concepts in Research—Week 5

Methods/Ethics/Theory Readings

Connaway and Powell, Chapter 3

Summative evaluation—is used in quantitative research to measure outcomes of a program.

Formative evaluation—is process evaluation, usually in qualitative research that looks for ways to improve the program that is being studied.

Connaway and Powell, Chapter 4

Snowball sample—a type of accidental sampling. If members of a certain population are difficult to identify, a researcher starts with the people they know that fall into the population the research is concentrated on, and have those individuals help locate other individuals of the same population. It is used in exploratory research and is considered a nonprobability sampling.

Simple Random Sample (SRS)—every member of a population has an equal chance of being chosen for a sample. If more than one draw of the population is made, the individuals who were chosen first need to be replaced in order for the random chance of being drawn to be the same. This however, is not generally practical, and if the same members are put back in the population from which the sample is taken, the chances are that some of the same individuals will be chosen. Mathematical formulas can be used to counteract this problem.

Wildemuth, Chapter 6

Middle-range theories—Imagine that theories are on a continuum from very specific to highly abstract and applicable to many phenomena. The middle-range theories fall in the middle of that continuum. They are specific enough to apply to a certain field, yet abstract enough that the phenomena could apply in a context of another field. For instance, if a research studied the use of libraries by online students it could also cross-over to the quality of education received by those students.

Grounded theory—Research in which the theory is established during and after the collection of data. Used in qualitative research, it allows the researcher to gather raw data, search for common themes and begin establishing a theory from these beginning concepts. The researcher then returns to the data and continue flushing out possible connections and theories. The memos that the research makes are then brought together to find a unified theory.

References

Connaway, L. & Powell, R. (2010). Basic research methods for librarians. 5th ed. Santa Barbara, CA: Libraries Unlimited.

Wildemuth, B. M. (2009). Applications of Social Research Methods to Questions in Information and Library Science. Westport, CN: Libraries Unlimited.

Week 4--Literature Review Readings

Key Concepts in Research—Week 4

Literature Review Reading

The Writer’s Handbook

What is the purpose of a literature review?

A literature review gives the researcher the opportunity to analyze a specific published body of work on the topic of concern. The researcher can summarize, classify, an compare studies, reviews, and theory related to the topic.

What are the basic parts when writing a literature review?

A literature review begins with an introduction so that the reader understands the topic, issue, or concern and the context within which you will be reviewing the literature. The body of the review summarizes the information you have gathered. The review should end with a strong conclusion that summarizes the major reviews that you mentioned in the body, points out strong relationships or flaws in the research, and concludes with discernment of the connections between the topic, the literature, and other current studies.

The Literature Review: A few tips on conducting it

What skills can you showcase by writing a literature review?

By conducting a thorough review of the literature, you can demonstrate your skills of information seeking and critical appraisal.

What questions should I ask myself when writing a literature review?

What research question, problem, or thesis does my literature review seek to define? Is this type of literature review interested in theory, methodology, policy, quantitative research, or qualitative research? What is the scope of the review? What type of publications should I use? Have I done a thorough job in information seeking? Has my search been wide enough to find the relevant literature, yet narrow enough to include only relevant material? Do I have enough sources for the length of my paper? Did I critically analyze the literature? Did I cite and discuss studies conflicting with my opinion? Is the review pertinent, appropriate, and beneficial?

Write a Literature Review

What are the stages of developing literature review?

According this article, the four stages of developing the literature review include: problem formulation, literature search, data evaluation, and analysis and interpretation.

How should I assess the literature?

Provenance—what are the author’s credentials? Objectivity—is the article fair, objective, or prejudicial? Persuasiveness—is the author convincing in the thesis? Value—does the work contribute to an understanding of the subject and were the author’s arguments and inferences convincing?

References

The Writer’s Handbook. http://writing.wisc.edu/Handbook/ReviewofLiterature.html

The Literature Review: A few tips on conducting it. www.writing.utoronto.ca

Write a Literature Review. http://library.ucsc.edu/print/help/howto/write-a-literature-review

Post Script to Article Review #1

The research I plan to conduct concerns the access of library resources to online/distance students.

Week 3--Developing the Study

Key Concepts in Research—Week 3

Developing the Study

Topic 14—Patten

Strict replication—repetition of an original study with the expectation of the same results.

Modified replication—conducting a study similar to the original study while making some major modifications such as a different population or using some other type of testing instrument.

Topic 15—Patten

What are the three major sources for locating literature electronically for the purpose of a literature review of previous research? 1) Sociofile for Sociological Abstracts and Social Planning/Policy & Development Abstracts; 2) PsycLIT for Psychological Abstracts; and 3)ERIC for education related abstracts.

What are some helpful tips for conducting a search? Use a Thesaurus to find proper descriptors of your topic, and use Boolean search operators, such as, AND, OR, and NOT.

Topic16—Patten

Conceptual definitions—defining the study in terms of the discipline and the theory. Words can mean different things in different settings and need to be clearly defined in relation to the research you will be conducting.

How do you show the importance of your topic? By conducting a literature review and showing that other researchers also found the topic important and conducted research on the topic.

Appendix D—Pryczak

Stratified random sampling—a method in which the population is divided into sub-populations and then randomly sampled. The method is used to produce more exact replication of the population as a whole.

Temporal stability—in research, temporal stability refers to consistent results using a certain testing instrument over time and across multiple forms.

Chapter 2—Wildemuth

Symmetry of potential outcomes—whether the results of your study confirm or disprove your hypotheses, the study will be useful.

Problem statement—A narrowly drafted research statement or question upon which your research will be based. It should be clear, specific, answerable, connected to significant concepts and phenomena and be relevant to the field.

Chapter 3—Wildemuth

Practice-based question—(In relation to librarians)Stepping outside your own personal experience, possibly teaming up with professionals from other libraries or other fields, such as education or psychology, to expand the knowledge within the field of library and information science.

Evidence-based information practice—the current standard of most information professionals to base decision making for their institution by identifying and using the assumptions of previous studies.

Chapter 5—Wildemuth

Null hypothesis—when in a hypothesis(… there is no relationship between the variables or no difference between one thing or another” (34).

Information foraging theory—suggests that people will use the nearest point of reference to decide whether to pursue the information they are seeking.

Chapter 3—Williamson

Research hypothesis—a more specific statement of the problem, but still too general for an operational hypothesis.

Operational hypothesis—a concise statement of the research hypothesis, developed to specifically guide the research study.

References

Patten, (2000). Topics in Educational Research. Pyczak Publishing.

Pryczak, F. (2008). Title Unknown. 4th ed. CA: Pyrczak Publishing.

Wildemuth, B. M. (2009) Applications of Social Methods to Question in Information and Library Science. Westport, CT: Libraries Unlimited.

Williamson, K. (2000). Research methods for students and professionals: Information management and systems. Wagga Wagga, NSW.