; DATA ANALYSIS
Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

DATA ANALYSIS

VIEWS: 3 PAGES: 2

  • pg 1
									Security, Controls,              Database (Data Repository)     Analysis, Interpretation,
Accuracy, Privacy                Accounting, Production, etc. Decision Making
DATA ANALYSIS, INTERPRETATION, DECISION PROJECTS
Planning
        Clear Objective or Question—Be Specific
        Why Important?
        Identify Data that can meet objective or answer question (population)
        Methodology—What tests or modeling will answer the question or meet the
        objective
Data Access
        Source or data, (Database or data repository)
        How to locate, what form is data in, and how to transfer
        If a Sample—Is the data an unbiased sample of the population?
Data Integrity Verification (Security, Controls, Accuracy)
        Is the data accurate and reliable?
        Any missing data? Decide how to handle missing data
        Data preparation—Get data in appropriate form for Data Analysis
Descriptive Data Analysis
        Tests: Statistical, Common Sense
        Outliers—How may they influence analysis & results?
        Report Results
                Do results answer the question or meet the objective defined in planning?
                Are results reasonable? (common sense)
        Conclusions
                Do conclusions really follow from the results? (Do results support
                conclusion?)
                Be careful about implications of cause and effect
                State Limitations & Assumptions
Predictive Data Analysis: Linear Regression, nonlinear regression, neural networks, etc.
        Descriptive Data Analysis helps researcher understand the underlying data
        Build model from past experience of data to predict future outcomes
                Use independent or predictive variables to predict the dependent or
                outcome variable(s)
                Usually have a model building sample, a testing sample (data sample used
                to give feedback during model building), and a holdout sample
        Report ting Results
                Do results answer the question or meet the objective defined in planning?
                Are results reasonable? (common sense)
                Is the model generalizeable to the holdout sample?
                        (the data not used to build the model
        Conclusions
                Do conclusions really follow from the results? (Do results support
                conclusion)
                Be careful about implications of cause and effect
                State Limitations & Assumptions

								
To top