Process of Quantitative Data Analysis and Interpretation
Process of Quantitative Data Analysis and Interpretation
of
Quantitative Data
Analysis and
Interpretation
Mgt E 521 ‐ Advance Practices and Quantitative Research to
Engineering Project Management
Presented by: Joseph Noel O. Coscos
FLOW OF TASKS IN ANALYZING QUANTITATIVE DATA
Inspect data for
Pre‐analysis Log in, check and
Select a software
package for Code data
Enter data onto
computer file and
outliers/wild
Clean data
Create and
document an
edit raw data codes,
Phase analysis verify
irregularities
analysis file
Preliminary Perform needed
transformations
Address missing
Construct scales,
composite
Perform other
actions values problems peripheral analysis
and recodes indexes
Principal Perform
descriptive
Perform bivariate
inferential
Perform
Multivariate
Perform needed
post hoc test
Analysis statistical analysis statistical analysis analysis
Perform
Interpretative Integrate and
synthesize
supplementary
interpretative
phase analysis
analysis
Pre‐analysis Phase
Select a software package It involves selecting statistical software for doing the data analysis. The two widely used
for analysis statistical software packages are the Statistical Package for the Social Sciences (SSPS now
called PASW or Predictive Analytics SoftWare) and Statistical Analysis System (SAS).
Coding is the process of transforming data into symbols, usually numbers. Certain variables
Code data
are inherently quantitative (e.g., age, body temperature) and may not require coding, unless
the data are gathered in categories (e.g., younger than 30 years of age versus 30 or older).
Enter data onto computer Coded data typically are transferred onto a data file via keyboard entry, but other options (ex.
file and verify Scanning of forms) are also available. Data entry is prone to error, so it is essential to verify
entries and correct mistakes.
Inspect data for outliers/wild Data cleaning involves two types of checks. The first is a check for outliers and wild codes.
codes, irregularities Outliers are values that lie outside the normal range. Wild code is, a code that is not possible.
A second data‐cleaning procedure involves consistency checks, which focus on internal data
consistency.
Clean data
The decisions that researchers make about coding and variable naming should be fully
documented. Documentation primarily involves preparing a codebook. A codebook is
Create and document an essentially a listing of each variable together with information about placement in the file,
analysis file codes associated with the values of the variable, and other basic information.
Sample SPSS Data file Sample Codebook
Preliminary
assessments
Assess missing values There are three missing values patterns. (1) Missing Completely At Random (MCAR), which
problems occurs when cases with missing values are just a random subsample of all cases in the sample.
(2) Missing At Random (MAR) which occurs if missingness is related to other variables but
not related to the value of the variable that has the missing values. And (3) Missing Not At
Random (MNAR), a pattern in which the value of the variable that is missing is related to its
missingness .
The distribution of data values for key variables also should be examined to assess any
Assess data quality anomalies, such as limited variability, extreme skewness, or the presence of ceiling or floor
effects.
Researchers often undertake preliminary analyses to assess biases, including the following:
Assess bias Nonresponse (volunteer) bias, Selection bias, and Attrition bias.
Assess assumptions for Most statistical tests are based on a number of assumptions, conditions that are presumed to
inferential test be true and, when violated, can lead to erroneous conclusions.
Missing Completely
At Random (MCAR)
Solution: Deletion
Missing At Random
(MAR)
Solution: Imputation (mean substitution)
Missing Not At
Random (MNAR)
Solution: Improved dataset/find data
Preliminary actions
Raw data often need to be modified or transformed before hypotheses can be tested.
Perform needed
Various data transformations includes: item reversals, count, recodes.
transformations and recodes
An index is a way of compiling one score from a variety of questions or statements that
Construct scales, composite
indexes represents a belief, feeling, or attitude. Scales, on the other hand, measure levels of intensity
at the variable level.
Perform other peripheral
Depending on the study, additional peripheral analyses may be needed before proceeding to
analysis substantive analyses. (ex. Data Pooling, Testing Cohort Effects, Testing Ordering (Carryover)
Effects)
Principal Analysis
Perform descriptive statistical Descriptive statistics. Analysis is conducted to describe the obtained data.
analysis
Perform bivariate inferential Inferential statistics. Analysis is conducted to make inferences about a population using
statistical analysis data obtained from the sample.
Multivariate analysis (MVA) is a Statistical procedure for analysis of data involving more
Perform Multivariate analysis than one type of measurement or observation.
Post hoc (Latin, meaning “after this”) means to analyze the results of your experimental
Perform needed post hoc test data.
Interpretative phase
Integrate and It is a stage in the systematic review process where extracted data (findings of individual
synthesize analysis studies) are combined and evaluated. The synthesis part of a systematic review will
determine the outcomes of the study.
Perform supplementary The analysis of research data provides the results of the study. These results need to be
interpretative analysis evaluated and interpreted, giving thought to the aims of the study, its theoretical basis, the
body of related research evidence, and limitations of the adopted research methods.
Interpretations of statistical results form the basis for the Discussion section of quantitative
research reports.
The interpretation of quantitative research results (the outcomes of the statistical analyses)
typically involves consideration of: (1) the credibility of the results, (2) precision of the
results, (3) magnitude of effects and importance, (4) underlying meaning of the results, (5)
generalizability of results, and (6) implications for the research for future research, theory
development, and practice.