Research Design, Data Collection Techniques and Selection of Subjects
Once the problem has been carefully defined, the researcher needs to establish the plan that will outline the investigation to be carried out. The research design indicates the steps that will be taken and in what sequence they occur.
There are two main types of research design:
1. Exploratory research
2. Conclusive research itself subdivided into
· Descriptive research and
· Causal research
Each of these types of research design can rely on one or more data collection techniques:
- Primary research
- Observation technique
- Direct communication with subjects, e.g. survey technique , interview or projective methods
- Secondary research , which essentially means reviewing literature and data sources, collected for some other purpose than the study at hand.
Irrespective of the data collection technique used, it is critical that the researcher analyze it for its validity and reliability .
Another critical consideration in determining a study’s methodology is selection of subjects. If the researcher decides to study all elements within a population, s/he is in fact conducting a census. Although this may be ideal, it may not be very practical and can be far too costly. The alternative is to select a sample from the population. If chosen correctly, it is considered to be representative of the population. In this case, we are dealing with one of the probability sampling techniques. If the sample is not representative, then one of the non-probability sampling techniques was employed.
When research is written up as a part of a newspaper article, there should always be an indication as to the methodology employed, as is the case with the attached article.
Session 2
ASA Citation Style
http://www.calstatela.edu/library/bi/rsalina/asa.styleguide.html
.How to Use Library Resources --information Tutorial by Healey Library -- good for narrowing down your research topics
http://www.lib.umb.edu/webtutorial/index.html
Why Conduct Research -- this is not sociology page but it is pretty brief and hits the point.
http://www.ryerson.ca/~mjoppe/ResearchProcess/WhyConductResearch.htm
research process - Go to Content -->go thrhough "Foundations"
http://trochim.human.cornell.edu/kb/index.htm
Session 3
Levels of measurement
http://www.ac.wwu.edu/~sngynan/slx7.html
Alternative explanation that treats binary variables differently
tp://ftp.sas.com/pub/neural/measurement.faq
Session 4
Measurement Tutorial
http://trochim.human.cornell.edu/tutorial/TUTORIAL.HTM
Measurement 2
http://www.seweb.uci.edu/users/cddooley/Text%20Links/WebSites.html
Using Survey Questionnaire Items to Measure Variables
http://www.css.edu/USERS/dswenson/web/ONLINE.HTM
Session 5
Survey Research Links
http://www.srl.uic.edu/srllink/srllink.htm
Session 6
Research Design page -- Go to Data Collection
http://members.tripod.com/~frede_dast/conseil1_a.html
Session 7
Survey Research Links
http://www.srl.uic.edu/srllink/srllink.htm
Session 8
Random Sampling
http://www.randomizer.org/
Experiment Design –Explanation
http://ericae.net/ft/tamu/Expdes.HTM
Session 9
Experiment -- Threats to Internal Validity
http://www.css.edu/users/dswenson/web/DESIGN.HTM
Field Experiment
http://www.usi.edu/libarts/socio/stats/index.html
Session 10
Qualitative Report -- Web pages, papers, texts, journals, & syllabi
http://www.nova.edu/ssss/QR/qualres.html
Content Analysis
http://academic.csuohio.edu/kneuendorf/content/
Session 12
On-line Stats Notes ( explanation)
http://www2.chass.ncsu.edu/garson/pa765/statnote.htm
Various Data Sources
http://www.trinity.edu/~mkearl/data.html
**********
Spss introduction
http://annick.stir.ac.uk/docs/spss/spss.html
Web Sites Recommended in Social Research Methods
2001
HERE ARE some recommended Method WEBSITES MENTIONED. BECAUSE WEB SITES SOMETIMES DISAPPEAR OR CHANGE ADDRESSES, THE URLS sometimes MAY RETURN THE "REQUEST NOT FOUND (404)" MESSAGE. THIS DOCUMENT PROVIDES A MORE UPDATED VERSION OF THESE LISTINGS AS OF September 2003, OMITTING DEFUNCT SITES, CORRECTING CHANGED ADDRESSES AND ADDING SOME NEWER LINKS. BUT ALSO KEEP IN MIND THAT THIS REVISION MAY T SOON BE OUT OF DATE TOO. STUDENTS SHOULD BE PREPARED TO LOOK FOR ALTERNATIVE SITES USING AN APPROPRIATE SEARCH ENGINE.
Chapter 1
The Logic of Social Research: Ruling Out Rival Hypotheses
Trochim's site at Cornell: http://trochim.human.cornell.edu/kb/index.htm. XXX
.
Chapter 2
Ethics: Protecting Human Subjects and Research Integrity
1. Ethics Standards by Professional associations of social researchers
American Psychological Association, http://www.apa.org/ethics/code.html, American Sociological Association, http://www.asanet.org/ American Anthropological Association, http://www.ameranthassn.org/. For links to many other professional ethics codes see http://www.ethics.ubc.ca/resources/professional/.
2. The federal government has taken an active role in protecting both the subjects of social research and the integrity of the research process. The Office for Protection from Research Risks (OPRR) of the National Institute of Health (HIH) has been renamed the Office of Human Research Protections (OHRP), and it publishes the relevant human subject guidelines at http://ohrp.osophs.dhhs.gov/polasur.htm. The current Code of Federal Regulations Title 45, Part 46 is quite detailed. But you can get a self-paced electronic tutorial on human subject protection at http://tutorials.rgs.uci.edu/. The Office of Research Integrity of the Department of Health and Human Services oversees investigations of scientific misconduct and provides links to the relevant statutes and regulations at its site: http://ori.dhhs.gov/. Especially interesting are that office's summaries of dozens of cases, some resulting in official judgment of misconduct and others closed without such a finding.
Chapter 3
Finding , Using, and Writing Research Reports: Library Usage and Report Style
1. In addition to searching libraries and journal databases, the internet can provide access to all the information on the World Wide Web (WWW). For an overview of the web and its role in social research see Appendix A. Psychwatch provides a listing of online journals in the psychology area, some of which are full text (http://www.psychwatch.com/journalpage.htm)XXX. Numerous search engines exist to help you explore the web, and they are reviewed in Search Engine Watch (http://searchenginewatch.com/). Because of the web's enormous size, you may need to focus your WWW searches. You can do this by using Boolean operators as described earlier in this chapter, often accessible through the advanced, power, or help options on a search engines menu.
2. Writing clear research reports requires extensive training and feedback. But the web offers several sites that provide helpful guidance. The Research Room (http://www.esc.edu/htmlpages/writer/menu.htm) offers instruction on each step of the report writing process. Paradigm is an online writing assistant that deals with all kinds of essays. The Purdue Online Writing Lab or OWL (http://owl.english.purdue.edu/) includes links with style information for such standards as psychology (APA) and the Modern Language Association (MLA). For a detailed online guide to APA format see Psychology with Style (http://www.uwsp.edu/psych/apa4b.htm).
Chapter 4
Theory: Tentative Explanations
1. Some web sites provide links to a variety of theories within particular social science disciplines. For theories in psychology and mental health see http://www.geocities.com/CollegePark/8639/psych.html.
2. The practical value of theory is illustrated by theories that help guide interventions as in the promotion of public health: http://rex.nci.nih.gov/NCI_Pub_Interface/Theory_at_glance/HOME.html
Chapter 5
Measurement Theory: Toward Validity and Reliability
1. The American Psychological Association posts information on testing and assessment at http://www.apa.org/science/testing.html. For a summary of tips in evaluating a measure see Test Evaluation: http://ericae.net/seltips.txt. For more on measurement theory especially in regard to level of measurement and appropriate statistics see Measurement Theory: Frequently Asked Questions: ftp://ftp.sas.com/pub/neural/measurement.html.
2. For some tests measuring constructs such as depression, locus of control, or self-esteem along with information about their reliability and validity see http://www.queendom.com/tests.html.
Chapter 6
Types of Measures: Finding and Using them
1. the Educational Resources Information Center's Clearinghouse on Assessment and Evaluation: http://ericae.net/. From this home page, you can find the Test Locator with links to reviews of tests and listings of published and unpublished tests: http://ericae.net/testcol.htm. For links to on-line tests visit: http://dir.yahoo.com/Social_Science/Psychology/Branches/Personality/Online_Tests/.
2. The web offers a variety of archival data sets. The Inter-university Consortium for Political and Social Research at the University of Michigan (ICPSR) describes itself as the world's largest archive of computerized social science data: http://www.icpsr.umich.edu/. Access to these holding is restricted to faculty and students of ICPSR member school. But everyone can visit other archives such as the United States Historical Census Data Browser with data for each state and county of the U.S. from 1790 to 1970: http://fisher.lib.Virginia.EDU/census/.
Chapter 7
Survey Data Collection: Issues and Methods in Sample Surveys
Resources on Internet Survey Methodology at the MIT web site: http://www.ai.mit.edu/projects/iiip/conferences/survey96/resources.html.
2. Using Existing Surveys -- Survey Research Center based at the University of Michigan (http://www.isr.umich.edu/src/)
the National Opinion Research Center based at the University of Chicago (http://www.norc.uchicago.edu/).
The U.S. Census Bureau conducts the most extensive surveys including not only the census but also the American Community Survey mentioned earlier in this chapter: http://www.census.gov/.
To help find survey data on the internet try the index provided by the University of California, San Diego: http://odwin.ucsd.edu/idata/.
Chapter 8
Inferential Statistics: Drawing Valid Conclusions from Samples
1. Power analysis can be used in the planning stage of research to estimate the sample size needed to achieve statistical significance for a given effect size. For a web-based sample size estimator see http://www.surveysystem.com/sscalc.htm.
2. There are many helpful and interesting statistical sites on the web (also see the web site suggestions in Appendix B). Clay Helberg’s Statistics on the Web is a good starting point with a wide range of links: http://www.execpc.com/~helberg/statistics.html. For interactive graphical illustrations of statistical ideas, see the Duke University listing of Java Applets: http://www.isds.duke.edu/sites/java.html.
Chapter 9
Designing Research For Internal Validity
1. Experiment – Info about Stanford prison experiment (co-designed by principal investigator Philip Zimbardo), which includes audio-visual material on this study: http://www.prisonexp.org/.
2. A set of web tutorials on topics in experimental research design has been posted on Bill Trochim's methods site: http://trochim.human.cornell.edu/tutorial/TUTORIAL.HTM.
3. One of the best ways to practice your understanding of research design is to analyze various studies—identifying the cause and effect variables and deciphering the research approach. A handy source of brief research reports is the National Clearinghouse of Undergraduate Research: http://clearinghouse.mwsc.edu/ (see the manuscripts nominated as outstanding).
Chapter 10
True Experimentation: External Validity and Experimental Construct Validity
1. Web experiments, the pros and cons of web experiments, and links to web labs can be found at http://www.psych.unizh.ch/genpsy/Ulf/Lab/WebExpPsyLabAnim.html. More web experiments and information for accessing or adding to the experiments' data files appear at http://www.olemiss.edu/psychexps/.
3. For web sites on experimental ethics and various nonconsensual human experiments, see http://www.dc.peachnet.edu/~shale/humanities/composition/assignments/experiment.html
4. For an extensive description of the Milgram obedience experiments visit this site: http://sociology.about.com/science/sociology/library/weekly/aa012501a.htm.
Quasi-experimentation: When Multiple Groups and Random Assignment Are Not Possible
1. For more on quasi-experimental design, see the chapter on this topic in Bill Trochim's electronic text: http://trochim.human.cornell.edu/kb/quasiexp.htm and its special coverage of such special designs as regression-discontinuity and regression point displacement. Also see his thoughtful reflections on the whole area of quasi-experimentation: http://trochim.human.cornell.edu/kb/advquasi.htm.
2. Single subject designs utilize multiple experimental treatments (e.g. reinforcements) and observations to draw causal conclusions. This special type of quasi-experimentation is frequently used in studies reported in the Journal of Experimental Analysis of Behavior. That journal's web site posts links to movies made of some of their published studies: http://www.envmed.rochester.edu/wwwvgl/jeab_movies/jeab_movies.htm. Another journal that often publishes single subject design studies is the Journal of Applied Behavioral Analysis. See its cite for numerous links to its own articles and to other journals on related topics: http://www.envmed.rochester.edu/wwwrap/behavior/jaba/jabahome.htm.
Statistical Analysis
Correlational Methods: Controlling Rival Explanations Statistically
1.Although conventional wisdom holds that researchers may not infer individual-level conclusions from aggregate-level data, a method has been proposed to do just that (King, 1997). You can find software for carrying out the necessary analyses at http://Gking.Harvard.Edu/stats.shtml. Other researchers are interested not in interpreting research from one level to another but rather in collecting and analyzing data at two or more levels in the same study. You can find student software for such hierarchical linear modeling at http://www.ssicentral.com/other/hlmstu.htm.
2. Because correlational designs risk the threat of spuriousness, they usually require multivariate analytic procedures to measure the association of interest after adjusting for possible confounding variables. The method called structural equation modeling (SEM) has become popular as a general approach for analyzing a wide range of correlational designs. For numerous links to related sites see Ed Rigdon's SEM page: http://www.gsu.edu/~mkteer/. For user-friendly software that performs SEM see the student version of AMOS on the SmallWaters site: http://www.smallwaters.com/amos/.
Chapter 13
Qualitative Research: Participant Observation
1. George Boeree developed for his qualitative research methods course: http://www.ship.edu/~cgboeree/qualmeth.html
2. There are numerous web sites providing resources for qualitative researchers including links on methods, philosophy, data, calls for papers, e-journals, and qualitative data analytic software among others. Judy Norris' QualPage lists many of these resources: http://www.ualberta.ca/~jrnorris/qual.html. Anthropology is the research profession that makes most use of qualitative methods, and the American Anthropological Association provides a comprehensive listing of research resources called Anthropology Resources on the Internet: http://www.ameranthassn.org/resinet.htm.
Chapter 14
Interpreting Research: Overview of Research Design and Review Methods
1. Find a literature review on the web using electronic indexing or abstracting services such as PsycINFO. As one example see Kirsch, I., & Sapirstein, G. (1998). Listening to Prozac but hearing placebo: A meta-analysis of antidepressant medication. This is available in the e-journal Prevention and Treatment: http://journals.apa.org/prevention/
2. If you are interested conducting a quantitative review yourself, you can find out more about meta-analysis in general, the statistical procedures employed for this method, and many links to other related sites: http://www.mnsinc.com/solomon/MetaAnalysis.html. For an example of software for meta-analysis see Ralf Schwarzer’s program: http://www.fu-berlin.de/gesund/gesu_engl/meta_e.htm.
Chapter 15
Applied Social Research
1. The journal Evaluation Review specializes in program evaluation, and its web site includes tables of contents and abstracts for some recent volumes as well as its mission statement and submission guidelines: http://www.math.ucla.edu/~johnston/journals/er/
2. Visit the Health and Human Services Program Evaluation Database, a federal government search engine for finding program evaluation reports: http://www.os.dhhs.gov/search/prog_eval.html
Appendix A
Social Research and the Internet
1. To get more background information about how the web works, there are sites to visit where you can explore the internet-related pages (http://www.howstuffworks.com/web-page.htm or http://www.learnthenet.com/english/section/www.html). There is also more information on internet-related terminology (http://www.netlingo.com/).
2. For more information on web page creation, visit the following sites. There is a reference to html tags (http://developer.netscape.com/docs/manuals/htmlguid/contents.htm). You can find guidance on web style (http://info.med.yale.edu/caim/manual/).
Appendix B
Statistics Review
1. What statistic should be selected for a particular problem? See Bill Trochim's Guide to Selecting Statistics at Cornell University: http://trochim.human.cornell.edu/selstat/ssstart.htm
2. Want to learn more about statistics? See the following: Statistical Instruction Internet Palette at Arizona State University: http://research.ed.asu.edu/siip/, Web Interface for Statistics Education at Claremont Graduate School: http://www.grad.cgs.edu/wise/ (especially the statistical applets), and The Chance course: http://www.dartmouth.edu/~chance/.