Advanced Research Methods

Pre-Class #1

Prior to January 14^th  

Working with basic stats tools

Please go to http://www.youtube.com. Type the words “SPSS tutorial” into the search box. Run several of the resultant videos explaining to you how to use the basic tools on SPSS.  Please pay particular attention to the regression tutorials.  Next, start working with SPSS by loading a dataset into SPSS for further processing.  To find a dataset, go to the “Datasets/ SPSS-ready datasets from Watertree Press” under Course Resources in my Onedrive directory you have been invited to.  Choose one dataset from the choices here after reading the dataset descriptions in “AAA SPSS-ready datasets from Watertree Press.” Using this dataset in SPSS, run a regression of at least two IVs against 1 DV (choose your own IVs and DV). Your deliverable is the resulting output file with the file extension *.spv.  Please create the filename with your name.  Example: Straub runs a regression and names the SPSS output file Straub.spv. This is one deliverable for this assignment.

**You will also want to view how to run simple descriptive statistics like frequency distributions, means, standard deviations, etc.  Also how to run Spearman and Pearson correlations, t-tests, etc.

Reading: TAM Exercise

The set of lecture notes (“TAM Exercise”) introduces you quickly to a scientific model explaining IT use.  It contains a regression run empirically testing the model at the very end of the graphics (pages 7-8).  Please access the dataset “Dataset 1 TAM.sav” to run the regression of this model for yourself in SPSS.

Questions for Pre-Class #1:

Questions: Please answer the following questions about the SPSS output on pages 7-8 of “TAM Exercise.”  What is the dependent variable?  What are the predictors?  What does the column labelled “Sig.” refer to?  How many hypotheses can be tested with this one regression run?  What are these paths?  Are there any paths in TAM that are not being tested in this run?  How would you interpret these results, i.e., what statistical inferences can you draw about statistical significance and strength of linkages between 2 IVs and the 1 DV? 

[Hint: There are two inferences.  One is about statistical significance.  The other is about explained variances (i.e., effect sizes.]

Class #1

Friday, January 17^th

8:30-11:30am

12:15pm-1:15pm

The Scientific Method and Quantitative, Positivist Research vis-à-vis Qualitative Work [Many-to-many relationships between methods, data collection techniques, and data analysis tools]

The Research Process

Theoretical Adaptation in QPR

Review of Basic Statistics with Demos

Readings:

• Trochim, Ch. 1 (“Foundations of Research”)
• Straub, Gefen and Boudreau (2005)-All pdfs are in the QPR (Quantitative Positivist Research) file directory
• Myers’ “Qualitative Research.pdf” (2003) (latter also available at: http://www.qual.auckland.ac.nz/ )
• Bhattacherjee, Ch. 3 (“The Research Process”)
• Bhattacherjee, Ch. 4 (“Theories in Scientific Research”)

Web site exploration:

• Larsen et al.’s “Theories Used in IS Wiki” (http://istheory.byu.edu/wiki/Main_Page); follow some of these hyperlinks to see what is included in explaining a given theory).

Lectures:

• “What is Science?”
• “An Introduction to QPR”
• “Parsimony and More Elaborate Research Models”

Class #2

January 28^th at 7pm

Readings:

·         Bhattacherjee, Ch. 5 (“Research Design”)

·         Handout: “Revision Plan to SE”

Lectures:

·         “Theory, Internal Validity, and Endogeneity Threats”

·         “Validity Priorities”

·         “Type I and Type II Errors Simplified”

·         “External and Internal Validity…  p-values in the Era of Big Data”

Class #3

February 11^th at 7pm

Comparative QPR Approaches or Methodological Tradeoffs

Studies using archival data vis-à-vis field studies (using self-reports or other data sources) vis-à-vis lab experiments

Working with Archival Datasets and Econometric Modeling

Endogeneity and Reverse Causality  (aspects of internal validity)

Readings/Handouts:

• Stone, Ch. 7 (“Empirical Research Approaches”)
• Trochim, Chs. 8-9 – Design and Experiments
• Jarvenpaa (1991) – Second-hand data sources: pros and cons
• Handout: Pros and Cons of Secondary Data Sources
• Trochim, Ch. 7 (“Survey Research”; i.e., almost always this means studies using self-reports in data collection, either field studies or experiments, for example).  Surveys are not a method.  They are a way to collect data.  Please note this.

Lecture:

• “Stone Ch. 7 Matrix”
• “Econometric Modeling and Regression”

Readings:

• Goertzel (2002)

Scan:

• Example of econometrics paper: Chen et al. (2019)

Class #4

 February 25^th at 7pm

Measurement Methods and QPR Scientific Validation

Multi-Item Measures of Latent Constructs

Multi-Dimensional Constructs

Readings:

·         Trochim, Ch. 5 “Intro to Measurement”

·         Trochim, Chs. 5 & 6 (“Scales, Tests, and Indexes”)

·         Polites et al. (2012) [Caveat: Their example of a second order reflective-reflective multi-dimensional scale is not correct. Otherwise, a useful paper.]

·         Bhattacherjee, Ch. 6 (“Measurement of Constructs,” but only pages 43-48)

·         Straub et al. (2004) “Validation Guidelines”

Scan:

·         Example of instrument validation paper: Samppa et al. (2019)

Lectures:

·          “Scales”

·         “Instrument Validation “

Note: Terms like items, indicators, scale items, response items, observations, proxies, surrogates, or measures have very similar meansings. They are all terms for the data attributes and data values we capture when measuring a research construct. Research constructs are abstractions and, therefore, most often NOT observable. The most interesting constructs are, in fact, said to be “latent.”  That is, NOT observable.

Trust is a latent construct.  Yet quite real.  [Trust, e.g., leads to sales.] It has even been measured through neural activity in the brain. Age and time, though, especially chronological time, is in many people’s minds observable. But chronological time is seldom an interesting construct, for that reason (IMHO-DW Straub).  Marketing research studies people’s ages, certainly, but they are mostly interested in experiential ages, like cognitive age or emotional age.  These are much more interesting as theoretical predictors.

Handouts:

·         “Assessing Reliabilities and Validities”

Web site exploration:

•      Online scales available for your use: (http://managementscales.com); search to see examples of usable scales online in the management field; try “computer self-efficacy,” for instance

Class #5

March 10^th at 7pm  

Demos of factor analysis, construct validity (convergent and discriminant validity), reliability

Interpretation of structural model results

Handouts:

• “Using Stats Tools Demo Handout”
• “VIFs, Etc.”

Class #6

March 24^th at 7pm

Scan:

·         Example of data analysis in journal-submitted paper: Guo et al. (2020). There are two documents here. One is the manuscript. The other is the appendices, which has a lot of the scientific apparatus for validating the instrument.

Class #7

April 3^rd at 7pm  

External Validity and Sampling

External Validity, Meta-Analysis, and Cumulative Research Traditions

Assumptions of Statistical Testing; Testing for Them and Dealing with Violations

Non-parametric Statistics

Analytical (or Math) Modeling

Contingency Theory & Qualitative Comparative Analysis (QCA)

Readings:

·         Textbook Trochim, Ch. 4 (“Sampling”)

·         Im and Straub (2015)

Lectures:

·         “A Few Basic, Useful Nonparametric Tests”

·         “Sampling Basics”

·         “Polling Misses in US Presidential Elections”

Scan:

·         Example of analytical modeling (math modeling) paper: Jung et al. (2019)

·         Example of paper qualitative comparative analysis (QCA) paper: Lee et al. (2019)

Class #8

April 17^th at 7pm

Oral Presentations of Final Group Project

If time, more on advanced analytical tools (CB-SEM); Preparing Manuscripts for Publication