By Brett DiNovi, M.A., BCBA
bSci21 Contributing Writer
The sheer volume of data that’s being created and stored on a global level is almost inconceivable, and it just keeps multiplying. This information can be crucial to an organization’s success– yet only a small percentage of data is actually analyzed. What does that mean for businesses? How can they make better use of the raw information that flows into their organizations every day?
Behavior Analysts spend countless hours entering client data into spreadsheets and charting them for visual inspection to guide decision-making for the learner, however the behavior of the change agent (business owners, leaders, or clinicians) is rarely analyzed to the same level of detail. Organization-wide data analysis that is actually used to drive day-to-day leadership decisions is even rarer.
What if healthcare and human services agencies could use big data to decrease overtime usage, increase utilization of authorized service units that maximize profit and patient care, decrease employee turnover, and identify individual and organizational trends to swiftly act on them to reverse negative trends proactively?

Isen Scholze
Isen Scholze, previously a leader at Bank of America, has spent countless hours developing complex algorithms that grind and churn thousands of data sets of employee behaviors into simple visual displays that drive peak performance for the agency. Every Monday morning, Mr. Scholze briefs the leadership team at BDA by displaying the previous week’s performance at the organizational and individual level. One example of his proprietary algorithms that have saved hundreds of thousands of dollars for the agency is one that tracks hundreds of employees’ use of hours approved by the funding source for each client. These data are then used to provide feedback to the employee, notifying them if they are breaching over the authorized service units or if they are underutilizing service units, negatively impacting the client. His graphic display catches these potential financial and service-quality hardships before they can negatively impact the organization. It allows for each employee to “self-correct” and change his/her behavior before it leads to disaster.

Figure 1 displays data that are derived from one of Isen’s algorithms that, at the push of a button, summarizes every client’s pace of service units used and where they are relative to the percentage required to stay on pace.

Figure 2 displays data produced from an algorithm that identifies the weekly trends of employee performance that determines if they are on pace to breach over or under the authorized service units. Note that on Nov 28th the algorithm generated feedback for 3 clinicians and this resulted in an immediate appropriate response by the clinician decreasing his/her service units used so as to not exceed the authorized hours and saved the company thousands of dollars.
To summarize, big data and predictive analytics can be customized and used in ways to change performance proactively rather than reactively. If organizations using applied behavior analysis for behavior change of clients could apply those same principles to leadership decisions derived from big data, enormous increases in efficiency of business processes and quality of services could be the outcome. Customized data analysis and visual display of employee behavior can and have produced sustainable change in these areas at BDA in the following ways:
- Rapid analysis of pinpointed employee behaviors, such as breaching over approved service units, can immediately be addressed and change employee performance to save hundreds of thousands of dollars before it’s too late.
- Predictive analytics can help leaders spot employees trending toward at a rate that could negatively impact the appropriate service hours to the client and correct this to meet contractual obligations.
- Customized algorithms can organize employee performance and act as data curation tools when conducting performance appraisals; previous performance is captured and easily accessible for determining employee bonuses.
- Appropriate organization of big data using inductive statistics can infer laws of regression, nonlinear relationships, and causal effects to reveal relationships and dependencies that may predict outcomes. In other words, behavior is more predictable, therefore easier to identify and control
- Custom-built algorithms can identify errors in billing automatically rather than manual inspection which results in more opportunities for human error.
Do you have experience with big data? Let us know in the comments below, and remember to subscribe to bSci21 to receive the latest articles directly to your inbox!
Brett DiNovi, M.A., BCBA has the unique and distinguished experience of studying the principles of applied behavior analysis under the rigorous scrutiny of both Dr. Julie S. Vargas (formerly Skinner) and Dr. E.A. Vargas at West Virginia University’s internationally recognized program. For the past 26 years, Brett has used behavior analytic principles to create large scale change across school districts, Fortune 500 companies using principles of Organizational Behavior Management (OBM), and across individual learners. Brett has been a OBM consultant in Morgantown WV, an instructor at West Virginia University, a guest lecturer at numerous universities, a speaker on multiple Comcast Newsmakers TV programs, an expert witness in due process hearings, has publications in the Journal of Applied Behavior Analysis, and has been in in executive leadership positions across schools and residential programs nationwide. In addition to an award from South Jersey Biz Magazine for “Best Places to Work,” an award for “Best of Families” in Suburban Magazine, and the distinguished “Top Ranked U.S. Executives” award, Brett’s proudest accomplishment is being a role model and father for his daughter and two stepchildren (one of which has autism). Brett can be reached at [email protected]
Be the first to comment on "Achieving Peak Employee Performance with Big Data"