3971intro1

44 %
56 %
Information about 3971intro1
Entertainment

Published on November 19, 2007

Author: Spencer

Source: authorstream.com

LIS 397.1 Introduction to Research in Library and Information Science Basic Concepts of Research:  LIS 397.1 Introduction to Research in Library and Information Science Basic Concepts of Research R. E. Wyllys Copyright 2003 by R. E. Wyllys Last revised 2003 Jan 15 The Goal of LIS 397.1:  The Goal of LIS 397.1 To help you learn about problem-solving and certain problem-solving tools Hypothesis-testing approach: a powerful time-tested way to tackle problems The tools: Descriptive statistics Inferential statistics Survey techniques Course Objectives:  Course Objectives You will Be introduced to systematic inquiry, also known as formal problem-solving and scientific research Be introduced to descriptive and inferential statistics Be introduced to social-survey techniques Gain experience in planning a research effort Gain experience in critiquing research reports Scientific Research:  Scientific Research Traditional View of Scientific Research Observations lead to an IDEA Idea is investigated, leading to a HYPOTHESIS Hypothesis is tested, tentatively accepted Related hypotheses are tested and accepted A THEORY subsumes the hypotheses More tests and investigations lead to a SCIENTIFIC LAW that subsumes the theories While the traditional view of scientific research remains valuable, the modern view tends to emphasize research as a way of solving problems Scientific Research:  Scientific Research A Modern View for LIS Fundamental theories are needed, but are hard to establish Practical problems arise constantly, and deserve our best efforts to solve them One of the best guides to the use of practical tools for solving practical problems is the hypothesis-testing approach Hypothesis-Testing Approach:  Hypothesis-Testing Approach Hypotheses: statements of what you believe is true, hope is true, or fear is true Why use hypotheses? Turning hazy ideas into a formal hypothesis forces you to think carefully about what you are trying to say Hypothesis-Testing Approach:  Hypothesis-Testing Approach Hypotheses State a relationship among two or more variables May be stated in positive or negative terms Must be capable of being tested as to whether they are “true” or “false” Hypothesis-Testing Approach:  Hypothesis-Testing Approach When is a hypothesis “false”? When it is not consistent with observable reality When is a hypothesis possibly “true”? When it appears to be consistent with observable reality The falsity of a hypothesis can often be definitely established. The truth of a hypothesis is always tentative. Hypothesis-Testing Approach:  Hypothesis-Testing Approach What is the general outline for solving problems using a hypothesis (or hypotheses)? Get an idea Turn it into a hypothesis Gather relevant data, i.e., make observations Check the hypothesis against the observed reality If the hypothesis is consistent with the observed reality, it may be true; at least, it has a chance of being true. If the hypothesis is not consistent with the observed reality, it is clearly false. Hypothesis-Testing Approach:  Hypothesis-Testing Approach To formulate a hypothesis, you must provide statements of Testable relationships among variables Related assumptions Pertinent definitions Hypothesis-Testing Approach:  Hypothesis-Testing Approach What are variables? Problems involve many factors: some that cannot be changed, and others that can or might be changed You need to think about a problem’s factors and identify the major ones Of these major factors, some cannot be changed; these are constants Others of the major factors can or might be changed; these are the variables Hypothesis-Testing Approach:  Hypothesis-Testing Approach What do assumptions deal with? Think about a problem’s context. What are factors that may affect the problem but are unchangeable or only very slowly changeable? Such factors probably need to be included in statements of assumptions about the problem and its context Hypothesis-Testing Approach:  Hypothesis-Testing Approach Explicit definitions are needed. Definitions Should be operational: e.g., IQ vs. intelligence IQ is a score derived from any one of several tests purporting to measure general intelligence. Because it measures something (i.e., performance on a test) it can be used in relation to other measures Intelligence is one's overall ability to cope with one's environment, and is at most moderately related to one's IQ score. No one really knows how to measure general intelligence, and not much can be done in a precise fashion with merely qualitative terms like "bright," "dull," "musically gifted," etc. Can be functional: i.e., can be tailored to the problem being studied Should be specific: i.e., should identify the intended thing and distinguish it from similar things Hypothesis-Testing Approach:  Hypothesis-Testing Approach Definitions are essential Definitions are difficult; you must expect the difficulty, and work to make them clear Bertrand Russell (1872-1970), one of the leading philosophers and thinkers of the Twentieth Century, described the elusiveness of definitions in a delightful vignette: Bertrand Russell on Definitions:  Bertrand Russell on Definitions “But,” you may say, “none of this shakes my belief that 2 and 2 are 4.” You are quite right, except in marginal cases--and it is only in marginal cases that you are doubtful whether a certain animal is a dog or a certain length is less than a meter. Two must be two of something, and the proposition “2 and 2 are 4” is useless unless it can be applied. Two dogs and two dogs are certainly four dogs, but cases may arise in which you are doubtful whether two of them are dogs. “Well, at any rate there are four animals,” you may say. But there are microorganisms concerning which it is doubtful whether they are animals or plants. “Well then, living organisms,” you say. But there are things of which it is doubtful whether they are living organisms or not. You will be driven into saying: “Two entities and two entities are four entities.” When you have told me what you mean by “entity,” we will resume the argument. From: Russell, Bertrand. (1942). How to Become a Mathematician. Girard, KS: Haldeman-Julius. What Constitutes a Hypothesis?:  What Constitutes a Hypothesis? Which of the following are hypotheses?* Life currently exists on Mars On average, men are taller than women On average, men are smarter than women Drinking coffee raises your IQ temporarily Drinking scotch raises your IQ temporarily People who get up early in the morning are better people than those who sleep late *Hint: The matter boils down to the question of which of these statements can be, at least in principle, tested to see whether it is true or false. Please try to decide that question in your own mind for each of the statements before moving to the next slides to see my opinion on the matter. What Constitutes a Hypothesis?:  What Constitutes a Hypothesis? Comments on the possible hypotheses Life currently exists on Mars Life either exists there or not. This is a hypothesis that, probably within a few decades, will be proven true or false, at least with respect to life as we know it. On average, men are taller than women Height is easily measured. This is a hypothesis that can be proven true or false by carrying out the necessary measurements. In fact, it is known to be true, on the average, though obviously some women are taller than some men. What Constitutes a Hypothesis?:  What Constitutes a Hypothesis? Comments on the possible hypotheses (cont'd) On average, men are smarter than women What does "smarter" mean? There are numerous tests of various mental abilities. On some of these tests, men attain higher scores, on the average, than women; on others of these tests, women attain higher scores, on the average, than men. On many of the tests, men and women average essentially the same. No one knows how to define an acceptable amalgam of scores on the various tests that could be used as a measure of "smartness." Even if such an amalgam were established, it would surely omit numerous mental qualities that cannot be measured objectively but nevertheless are vital to a person's coping with his or her environment, which is what we intuitively mean by "smartness". Conclusion: Since the statement cannot be proven either true or false, it is not a hypothesis. What Constitutes a Hypothesis?:  What Constitutes a Hypothesis? Comments on the possible hypotheses (cont'd) Drinking coffee raises your IQ temporarily The statement refers to IQ, which means your score on one (or more) of several standard tests. It would be relatively easy to devise a test of whether you score higher, on the average, on an IQ test after consuming coffee. This is a hypothesis that can be proven true or false. Drinking scotch raises your IQ temporarily Just as with respect to the previous statement, a test could easily be devised to determine whether you score higher, on the average, on an IQ test after consuming scotch. This is a hypothesis that can be proven true or false. What Constitutes a Hypothesis?:  What Constitutes a Hypothesis? Comments on the possible hypotheses (cont'd) People who get up early in the morning are better people than those who sleep late The crux of this statement lies in what one means by "better people." I cannot imagine any reasonable definition of "betterness" that would be both widely agreed upon and also measurable objectively. Till someone comes up with such a definition, I have to hold that this statement is not a hypothesis that can be proven true or false. Examples of Hypotheses:  Examples of Hypotheses Some Possible LIS hypotheses Raising our library’s overdue fines will result in fewer books being returned overdue Patrons will ask more reference questions when librarians dress informally than when they dress formally The proportion of Library X users who like X’s online catalog menus is higher than the proportion of Library Y users who like Y’s online catalog menus. (“Experienced users prefer X’s catalog.”) Note: A light-hearted but sound discussion of using hypotheses in research can be found in: Pirsig, Robert M. Zen and the Art of Motorcycle Maintenance. New York, NY: Morrow; 1974. See, especially, Chap. 9. Two Types of Hypotheses:  Two Types of Hypotheses General Hypotheses Concern variables directly related to the problem being studied Statistical Hypotheses Are a subclass of general hypotheses Are tools Are used as parts of efforts to determine whether general hypotheses are true or false General Hypotheses:  General Hypotheses A general hypothesis Makes a statement about relationships believed, hoped, or feared to exist among variables Should say what you think the situation actually is Another Look at Types of Research:  Another Look at Types of Research Research Types that Emphasize Tests of Hypotheses Causal Comparative (“Ex Post Facto”) Research Correlational Research Historical Research (some, not all) Quasi-Experimental Research True Experimental Research Research without a Hypothesis-Testing Emphasis Action Studies Case Studies, Field Studies Descriptive Studies Developmental Research Historical Research (some, not all) Types of Experimental Research:  Types of Experimental Research True Experimental Research (TER) Investigates possible relationships by exposing one or more experimental groups to one or more treatment conditions and comparing the results to those of one or more control groups not receiving the treatment(s)1 Quasi-Experimental Research (QER) Approximates the conditions of a true experiment in a setting that does not allow the control and/or manipulation of all relevant variables1 Both TER and QER may, and should, be carried out using hypotheses 1Adapted from: Isaac, S.; Michael, W. B. Handbook in Research and Evaluation. 3rd ed. 1995 True Experimental Research:  True Experimental Research The investigator can control the treatment(s) given to the subjects, and can compare the effects observed on the subjects with the effects (or lack thereof) observed on a control or comparison group with different or no treatment(s) Examples of True Experimental Research:  Examples of True Experimental Research 1 Variable, 2 Treatments Investigator puts a container of pure water and a container of salt water into a refrigerator; then observes and compares the temperatures at which the containers freeze. 1 Variable, 4 Treatments Investigator prepares 5%, 10%, 20%, and 40% solutions of salt water and chills them; then observes and compares the different temperatures at which the various solutions freeze. 1 Variable, 2 Treatments Investigator administers, via double-blind technique, pills containing an experimental oral vaccine against Disease X to a group of healthy people (subjects) and pills containing a placebo to another group of healthy people (control group); then observes and compares numbers of subjects and controls who subsequently contract Disease X. Examples of True Experimental Research:  Examples of True Experimental Research 3 Variables, Many Treatments Investigator plants 3 varieties of wheat in scattered plots of ground in a field, and treats different plots with different amounts and kinds of fertilizer as well as different amounts and schedules of watering; then observes the various amounts of wheat harvested under the various combinations of conditions (i.e., treatments). Note: This kind of complex experimental situation is often required because of the long times involved in agricultural experiments. It is the situation for which the powerful statistical tool known as Analysis of Variance was invented. This tool has turned out to be valuable in many other areas of complex research. Examples of True Experimental Research:  Examples of True Experimental Research 2 Variables, Many Treatments Investigator compares effects of two methods of pupil evaluation by studying the performance of children in 23 elementary schools in a suburban school district, with children being chosen at random for evaluation by the two different methods1 1Adapted from: Isaac, S.; Michael, W. B. Handbook in Research and Evaluation. 3rd ed. 1995 Quasi-Experimental Research:  Quasi-Experimental Research The investigator cannot control the treatment given to the subjects, and hence cannot directly compare the effects observed on the subjects with the effects (or lack thereof) observed on a control or comparison group with different or no treatment(s) Instead, the investigator must try to identify different states of the subject (e.g., different initial conditions, and/or different post-treatment conditions) and connect these different states with different treatments, in a retroactive or after-the-fact fashion Quasi-Experimental Research:  Quasi-Experimental Research In other words, in quasi-experimental research, the investigator tries to approximate experimental research by identifying situations that have a factor of interest and comparing them with situations that are as similar as feasible but do not have that factor Quasi-Experimental Research:  Quasi-Experimental Research One common type of situation in QER: The Pseudo-Experiment Two reasonably different treatments have been applied to groups that have not been (and cannot be) selected in advance to be as much alike as possible; differing results are available for comparison, but initial conditions are hard to pin down accurately. Example: Four high-school foreign-language classes have used two different methods of practicing memorization: spaced practice and massed practice. Various differences may exist in the assignment of students to the four classes, in the teachers of the four classes, in the classrooms used, in the times of day when the classes are taught, etc. All of these differences may (or may not) affect the differences in success in memorization; the differences in success therefore may or may not be due to spaced practice vs. massed practice.1 1Adapted from: Isaac, S.; Michael, W. B. Handbook in Research and Evaluation. 3rd ed. 1995 Quasi-Experimental Research:  Quasi-Experimental Research Another common type of situation in QER: The Survey Information about a group of people is to be obtained via a survey. The survey should be designed so as to enable comparisons of Different conditions or levels of certain variables of interest (e.g., gender, age, education, income, occupation, political party identification); against Different conditions or levels of certain other variables of interest (e.g., political party affiliation, hobbies, crime rate in ZIPcode of residence, attitude toward charitable giving to various organizations, health) Analysis of the survey will try to link various variables (inputs, possible causes) with various other variables (outputs, outcomes) Quasi-Experimental Research:  Quasi-Experimental Research You should note that In both of these common situations in QER, the Pseudo-Experiment and the Survey, the statistical techniques of correlation analysis, analysis of variance, and regression (simple and multiple) can often yield useful results True Experimental Research--if it can be performed--will probably yield more definitive results than Quasi-Experimental Research. Distinctions between TER and QER are sometimes tenuous and difficult to draw Nevertheless, QER is an extremely important way of carrying out research in LIS Another Look at Some LIS Research Questions (Hypotheses):  Another Look at Some LIS Research Questions (Hypotheses) Hypothesis: Raising our library’s overdue fines will result in fewer books being returned overdue TER approach: Raise the fines by a trial amount and see what happens; perhaps try this several times with different amounts of increase QER approach: Do a survey of library users to ascertain (1) attitudes toward higher fines and (2) possible changes in behavior of users with respect to returning books on time and borrowing books at all What are the possible consequences of each of these approaches? Which approach would you choose if you were the library director? Another Look at Some LIS Research Questions (Hypotheses):  Another Look at Some LIS Research Questions (Hypotheses) Hypothesis: Patrons will ask more reference questions when librarians dress informally than when they dress formally TER approach: Over a period of at least several weeks, make a random choice of days on which librarians will dress formally and other days on which they dress informally, and see what happens QER approach: Do a survey of library users to ascertain (1) attitudes toward formal vs. informal dress and (2) possible changes in behavior by users toward librarians What are the possible consequences of each of these approaches? Which approach would you choose if you were the library director? Another Look at Some LIS Research Questions (Hypotheses):  Another Look at Some LIS Research Questions (Hypotheses) Hypothesis: The proportion of Library X users who like X’s online catalog menus is higher than the proportion of Library Y users who like Y’s online catalog menus. (“Experienced users prefer X’s catalog.”) TER approach: None, except possibly to change X’s menus to those of Y and note possible increase or decrease in use of the online catalog, then change Y’s menus to those of X and note changes in use; then try to figure out which change had greater impact on the users QER approach: Do a survey of users of X and users of Y to ascertain their ratings of the respective menus on a Likert scale What are the possible consequences of each of these approaches? Which approach would you choose if you were a library director setting up a new online catalog and wanting to choose between Type X and Type Y menus? Summary: Experimental Research:  Summary: Experimental Research Experimental Research exists in two varieties True Experimental Research Quasi-Experimental Research Both types of experimental research Can yield useful results Involve difficulties and possible pitfalls Are made more productive and precise through the use of carefully thought out and stated hypotheses Because LIS concerns the complex interactions of people and information-bearing entities (InBEs), many research issues in LIS are best studied by the techniques of QER A HYPOTHESIS TESTED!!! A RESEARCH QUESTION ANSWERED!! The light of knowledge will shine!:  A HYPOTHESIS TESTED!!! A RESEARCH QUESTION ANSWERED!! The light of knowledge will shine!

Add a comment

Related presentations

Related pages

bYTEBoss ArtBooks_HolList_Dec07

Sheet1. Cambridge University Press (2007), Edition: 1, Hardcover, 316 pages. Reaktion Books (2007), Hardcover, 142 pages. Routledge (2006), Edition: 1 ...
Read more

LIS 397.1 Introduction to Research in Library and ...

LIS 397.1 Introduction to Research in Library and Information Science Basic Concepts of Research R. E. Wyllys Copyright 2003 by R. E. Wyllys Last revised ...
Read more

bYTEBoss LISTADO_DE_PRIORIZADOS_VIGENCIA_20091_cimitarra

3971intro1; LIS 397.1 Introduction to Research in Library and; Revenue Outturn Appendix; Strategic Services. Social Services. ERH. Educ Cu; Ffmay26 2010
Read more