|
Data Management
- Data Coding
- Data Entry
- Data Transformation
- Data Validation & Cleaning
- Open Ended Coding
(for verbatims)
|
Data Analytics
- Frequency distribution
- Data Tabulation
- Tables and Charts
- Basic/Standard Analysis
- Advanced Analysis
- Statistical Modeling
|
Data Visualization
- Charting
- Report preparation
- Running Client-Proprietary Software
|
Data Processing – Preparation and Management
|
Statistical Data Management is a systematic process that transfers raw data to intelligent data and requires the application of an integrated process. Data Collected may be available in various formats: Excel, Access, ASCII, .dat etc. It may have been stored in a database linked to the survey or punched directly into a database. Data collected in various formats needs to be converted or imported. It may be required to be transferred to an SPSS or SAS application or as required by the development team for further analysis. Most of these applications require structures and this data in the process of migration may be corrupted.
The Process Involves:
Extracting and loading data into client specified database format
Cleaning, de-duplicating, parsing and de-parsing etc,
Coding Verbatims, Generating tables and graphs, sig-testing, cross tabulation
Data Cleaning and Validation
Skewed data is a nightmare. Not only does it take a lot of time it is the first step leading to wrong decision making. Data needs to be checked, validated and cleant before it is sent for further processing. It is not as simple as it seems. It requires an experienced hand to understand “the emotion in the data” to be able to provide data in its absolute truth.
Data Coding
Every response is coded for the simple reason that unless coded it cannot be used to report findings. Applications do not understand statements and requires coding to make sense of the data. However, we frequently require responses from the audience in statements. Although, applications today have modules that detects and codes verbatim comments it is most frequently done manually. We also come across the terms netting, bucketing, basket etc in this step which simply is the term used for structuring the data in a more meaningful or easy to use form.
Data Tabulation
The “coded” and now easy to translate data is tabulated or summed to provide the first useful insights. This is not enough to meet the objective of the Market Research need. The data tabulated cannot provide any correlation or cannot be merged with different data points/structures to provide meaningful information. We require processing the tabulated data through the next step: Data Analysis.
Read the complete Article on Outsourcing Data Processing to India
Data Analytics and Modeling
At a time when companies in many industries offer similar products and services and use comparable technologies, in variety of business problems, the use of predictive analysis is the remaining key process of differentiation. As a result, organizations that effectively implement this emerging use of extensive analytics and fact-based decision-making from the top down are staying ahead of their competition.
Standard analyses include regression, correlation, PCA, significant testing, etc.
Enhanced analyses are more goal-driven and include conjoint analysis, TURF analysis, GAP analysis, TREND analysis, etc.
What we can do for you?
Data Summarization - Cross Tabs, Correlation etc
Data Visualization - Bar Plots, Bubble Plots, Control Charts, Histograms, Pareto Charts, Pie Charts, Q-Q Plots etc
Comparing Samples - Binomial Tests, Fishers Tests, Chi-Square Tests etc
Regression Analysis - Linear, Ordinal, Log-Linear models, and Non-Linear Regression etc
ANOVA - Mixed Models, Multiple Comparisons, Random Effects etc
Nonparametric Tests - Signed Rank Tests, Chi-Square GOF, Non-Parametric Anova etc
Multivariate Analysis - Discriminant Analysis, Factor Analysis, Cluster Analysis etc
QC Charts and Analysis - Simple Control Charts, Process and Quality Control Analysis, Pareto Chart Analysis, Six Sigma DMAIC Tools etc
Survival Techniques - Parametric Survival Models, Cox Regression etc
Time Series - Auto Correlation, Seasonal Variation Analysis, Lag and Spectrum Plots, ARIMA etc
Data and Text Mining - Visualizations, CRISP-DM, Text and web Mining
Missing Value Analysis - MCAR, MAR, Multiple Imputation
Bootstrapping - Bootstrap Methods, Jack Knifing, Custom Simulations
Read the complete Article on Outsourcing Data Processing to India
Reporting
“A good cook is one who pays extra attention to how is creation is presented”. Absolutely right!
Preparing a presentation is definitely an art, is dealt with utmost attention to detail and is developed by an experienced designer. Don’t we wear our best suits on a Presentation day?
The Analytics insight is presented in a Power Point presentation and in most cases this determines customer delight. It is very important to understand the customer’s Style in presenting the report to ensure the results are appreciated. It is true in this case that the “first impression is the best”. If the Output presented matches the customers taste in terms of color and layout design you can rest assured you will get an incentive.
Reporting is typically based on client specified format and this can be done by requesting for a sample report template that they use as a standard. This sample could also be used to come up with a customized design and style that shares the same feel as well as retains the Agency or Company identity. Clients may request reports in more than one format - PPT, EXCEL (Back up file), MS Word etc.
|
