StrategyDriven

Evaluation control programs must be credible in order to add meaningful value to the organization. Credibility is built not only by the quality of the data collected but also by the method by which it is collected, how it is combined, and how it is interpreted to create useful information in support of decision-making.

Data synthesis or data combination should be performed in a manner that enhances the credibility of the conclusions drawn. This process is comprised of two important parts: data prioritization and data association.

Data Prioritization

Begin with the Most Credible Data - data having high accuracy or believability. Examples of data from high to low credibility include:

• what we see - observable, measureable
• what we read - what others observed, measured
• what we hear - what others tell us, opinions

Begin with Refined Data - data having human intelligence and analysis incorporated into the findings; often by individuals closer to or directly involved with the occurrence or possessing special skills and experience in data analysis. Examples of data from high to low refinement include:

• root cause analyses
• self assessment reports
• apparent cause analyses
• benchmarking comparison reports
• management observation reports
• organizational performance measures
• condition reports, operating logs and records
• surveys, opinion polls

Begin with the Most Significant Data - data gathered from impactful events; revealing true organizational values and common behaviors because people tend to return to their core convictions and primary habits during times of increased stress.

Begin with the Most Recent Data - data representing the way things are or what is rather than the way things were or what was. The time period representing recent data varies depending on the rate of change of the parameter of concern.

Data Association

Ensure Only Like Data is Combined – data having equivalent meaning, units, timeframes, and absolute references can be relatable. Combining data not possessing these qualities may lead to logic errors when making decisions including:

• Bad Analogy - claiming that two things are similar when they aren’t
• Extended Analogy - claiming that two things related to a third are therefore, by extension, related to each other
• Argument from Spurious Similarity - suggesting two items sharing some similar characteristics is evidence of a relationship between them
• Equivocation - asserting the existence of a relationship between multiple items by associating a word with more than one meaning to each differently; one with one meaning, another with a second meaning, and so on…

Ensure No Double Counting of Data - individual data points occurring more than once in a combined data set. This circumstance can occur when combining two or more refined data sources based on one or more common underlying facts and tends to result in over-stated conclusions.

Following these guidelines will enhance the accuracy and credibility of the information presented to decision-makers. Improved accuracy reduces decision risk while increased credibility enhances the confidence decision-makers have in their chosen direction. Thus, sound data synthesis, through proper data prioritization and association, enhances the organization’s opportunity for success.

Popularity: 30% [?]

If you're new to StrategyDriven, please consider subscribing to our RSS feed. Thanks for visiting!