Research Methodology Part 9: Testing of Hypotheses
(2007) Research Methodology Part 9: Testing of Hypotheses. Library Instructional Material.
Full text available as: |
Abstract
This presentation material in PowerPoint is the ninth of an eleven-part package designed and used regularly for teaching research methodology particularly to post-graduate students and research scholars of Library and Information Science. As mentioned earlier, it is very unfortunate that research in librarianship has very rarely resorted to experimental design and testing of hypothesis. A hypothesis, as a tentative hunch, explains the situation under observation so as to design the study to prove or disprove it. What a researcher is looking for is a working or positive hypothesis. It is very difficult, laborious and time consuming to make adequate discriminations in the complex interplay of facts without hypothesis. It gives definite point and direction to the study, prevents blind search and indiscriminate gathering of data and helps to delimit the field of inquiry. Absence of a clear theoretical framework and lack of ability to logically utilise it and failure to be acquainted with the available research techniques are the main reasons for not resorting to formulation and testing of hypothesis. It is also equally important that the researcher should not start out to prove or disprove hypothesis nor try to defend his hypothesis. A good and usable hypothesis should be precise, simple but not obvious, conceptually clear, have empirical referents, specific and limited in scope, consistent with most known facts, state relationship between variables and related to available techniques and body of theory. A statistical hypothesis is a predictive statement (usually put in the form of a null hypothesis and an alternate hypothesis) capable of being tested by scientific methods and that relates an independent variable to some dependent variables. What a researcher bets in advance of his experiment that the results will agree with his theory and cannot be accounted for by the chance variation involved in sampling is hypothesis testing. In other words, hypothesis testing is a procedure which enables researcher to decide whether to accept or reject hypothesis.
There are many concepts like Type-I and Type-II errors, one-tailed and two-tailed tests, significance test, important parametric tests like Z-test, t-test, F-test and non-parametric tests like Kolmogorov-Smirnov one sample test, Runs test for randomness, sign tests, Fisher-Irwin test, Mc Nemer test, median tests, Chi-square test, Wilcoxin-Mann-Whitney U-test, Wilcoxin matched pair or signed rank test, Kruskal-Wallis or H test, Kendall's co-efficient of concordance, etc. need to be understood for testing of hypotheses. They are yet to be employed effectively in library research. What is surprising is that user research is market research in librarianship but user research has not even adopted a fraction of methodology adopted in market research.
Part 1: http://eprints.rclis.org/archive/00009276/
Part 2: http://eprints.rclis.org/archive/00009533/
Part 3: http://eprints.rclis.org/archive/00010170/
Part 4: http://eprints.rclis.org/archive/00010879/
Part 5: http://eprints.rclis.org/archive/00010880/
Part 6: http://eprints.rclis.org/archive/00010881/
Part 7: http://eprints.rclis.org/archive/00011775/
Part 8: http://eprints.rclis.org/archive/00009083/
| Keywords: | research, LIS, research methodology, hypothesis, null hypothesis, alternate hypothesis, hypothesis testing |
|---|---|
| Subjects: | A. Theoretical and general aspects of libraries and information. > AZ. None of these, but in this section. |
| ID Code: | 12947 |
| Deposited By: | Sridhar, M S |
| Deposited On: | 11 March 2008 |
| All fields: | Show all fields |
Archive Staff Only: edit this record