Box Plot Outlier Data Analysis on Research Datasets

If you conduct online survey research - or do any type of data collection - analyze results in one of several box & whisker plot outlier data analysis templates. Some templates are free, others may be purchased with a Data Analysis Template Yearly Subscription or individually. Below are a few templates that might be of interest:

 

Excel Find Outliers Analysis AddIn

 

Use this Excel Find Outliers Analysis AddIn if you have a single set of research or survey observations and need to quickly analyze, flag and visualize them for unusually high or low values: these can skew the rest of your research analysis. You can then perform an Analysis of Variance (ANOVA) on the segments created with this AddIn.

 

Multivariate Observations Outliers Analysis

If your research involves data collection on more than measure, use a FREE Scorecard Outlier Analysis Template that calculates the Z-Score statistic for three separate numeric (eg. $, % or #) variables. The Scorecard variable is used to create an 'Outlier Flag' column variable: values highlighted in yellow indicate unusually high Z-Score values across the three numeric variables being analyzed. Also an excellent alternative to the K-Means or Hierarchical clustering algorithms commonly used for segmentation analytics: you save a lot of time finding and interpreting the optimal number of segments.

 

Scatterplot Correlations Between Variables

 

Find correlations between variables in your research dataset by creating a properly formatted scatterplot with 5 different trendlines (eg. linear regression, exponential smoothing etc.). Each trendline is plotted with its equation and co-efficient. This FREE Weekly Period Scatterplot and Trendlines Analysis Template is a great tool for visualizing and choosing continuous variables that might work well together in a forecasting or predictive model.

 

Analyze Observations Over Time for Trends

The TIME SERIES Outlier and Anomaly Detection Template will let you automatically analyze your research observations over 12 time periods (eg. days, week, months, years). It will apply outlier detection theory to find, flag and let you visualize hidden patterns and trends in time series data. Also an excellent alternative to the K-Means or Hierarchical clustering algorithms commonly used for segmentation analytics: you save a lot of time finding and interpreting the optimal number of segments.

 

Do Financial Ratio Analysis on Your Datasets

 

Your research might include data for assessing industry or company performance ratios over weeks, months or years that makes it suitable for the Advanced Ratio Analysis Toolkit. Copy and paste as many columns of your data into the grey-shaded rows 1, 2 and 3. When you click the "Ratio Analysis" button in the top right-hand corner of the worksheet, follow the prompts to build your own chart visualizing "Ratio Analysis", "Growth Rate" and "Market Share" trends in financial data. You will also see which ratios are unusually high or low (yellow shaded cells) values based on the outlier statistical theory of the box & whisker plot.

 

I hope that these templates help you to analyze and visualize research datasets.