Identifying metrics, measuring output and analyzing information are common practices for every business. Leveraging data to make the best decisions is key to building a successful organization. There’s no doubt data analytics is here to stay. However, did you know 80% of HR professionals score themselves low on their ability to analyze data? As a result, an opportunity to provide guidance and value can be completely lost.
After this course, you’ll understand the important terms and concepts associated with HR analytics and the purpose of developing them in your organization. Start using techniques to begin the analytics process. Learn a new skill that'll help you and your HR department gain credibility and make better decisions.
HR Jetpack is recognized by SHRM to offer Professional Development Credits (PDCs) for SHRM-CP or SHRM-SCP. This program is valid for 1.50 PDCs for the SHRM-CP or SHRM-SCP. For more information about certification or recertification, please visit shrmcertification.org.
This activity, has been approved for 1.50 HR (Business) recertification credit hours toward aPHR™, PHR®, PHRca®, SPHR®, GPHR®, PHRi™ and SPHRi™ recertification through HR Certification Institute® (HRCI®). For more information about certification or recertification, please visit the HR Certification Institute website at www.hrci.org.
The use of the HRCI seal confirms that this activity has met HR Certification Institute's® (HRCI®) criteria for recertification credit pre-approval.
Title: Predictive Analytics Demo
Module: Analytics - A Deeper Dive
Let's continue our journey by reviewing a simple example related to predictive analytics. In a previous lesson I mentioned correlation and regression analysis. Correlation is a good first step. It provides great insight into the relationships between variables helping to answer the question if one variable increases in value, what happens to the second variable? This starts to give us more insight into where we should focus our time, energy and money.
For example, I pulled together sample data for four typical business measurements including tenure, performance ratings, merit increase and employee labor productivity. Each row represents an employee.
Let’s make the following assumptions. Your employees tenure or time with the company ranges from 1 year to 35 years. Performance ratings are on a scale of 1-4, 1 being the lowest rating and 4 being the highest rating. An average merit increase is around 2.5% but can get as high as 5%. And finally, I've input a calculation for employee labor productivity. I have come from the world of manufacturing and to measure employee labor productivity we divide employee labor hours by the number of products produced by an employee during a time period. For this example, we used production data for one year. I assumed a 40 hour work week is in place and all the employees on this list are full-time. All of them have a similar job producing the same widget. I pulled in sample production data for each employee. This number in column D represents the personal productivity level in one week. I took the number of widgets produced by an employee which ranges from 300 to 2000 and divided by 40 to calculate the personal production level.
If we just want to calculate the correlation between two sets of data, we can use the correlation function. I’ll use tenure and performance rating. Let's go down to an empty cell. Type the = sign, CORREL, open parenthesis, select A2:A50, B2:B50, close parenthesis. At .78, this is telling us there is a correlation between tenure within the organization and the performance rating. This is called the correlation co-efficient. It appears that as tenure increases, the performance ratings are increasing as well. The closer the correlation coefficient, this number right here, is to 1, the more positively correlated the variables. If this were a real life scenario, I’d be a little concerned with how your short tenured employees are being rated.
Let’s look at another variable, a comparison between tenure and weekly production rate. How about putting this into a chart. Let's select our data in column A and column D. Go to the Insert tab, select scatter diagram. We use the scatter diagram because it displays the raw values of data – the X in this case is tenure and the Y is employee productivity. Let's update the title to read "Tenure – Productivity".
Go to Chart Design, Add Chart Element and select “Add trendline”. The steeper this trend line, the greater the relationship between two variables. If it moves from the upper left hand corner tilting downward to the right, it’s a negative correlation. If it's flat, there’s no correlation and if it extends from the upper right corner down towards the left corner, it’s a positive correlation. I wouldn’t necessarily recommend displaying this during a staff or leadership meeting. However, it’s a great way to look at the data for a quick assessment of the variable relationship.
Let’s take it a step further. Using the Data Analysis tool in excel, you can easily conduct more advanced calculations. Note, you may need to Add In this tool by going to Tools and searching for Add Ins. For excel 2016 on the Mac, go to Tools and select "Add Ins". For the PC, go to the file button, select Options and Add Ins.
Let’s move the cursor down, select the Data tab, go to Data Analysis, select Correlation, select Ok, select the entire range of data, starting from A1 to D1 down to D50. Be sure to Click labels in first row, select output on a separate sheet and hit okay. I'll clean up our table with a bit of formatting magic. This is telling us that tenure is positively correlated to ratings, getting a merit increase and labor productivity. It’s the same for ratings being correlated to the merit increase and productivity. And the merit increase being correlated to labor productivity. Naturally, each of the variables is perfectly correlated to itself shown by the set of ones going diagonally through the table.
Remember this is only sample data to help you understand the correlation concept and calculation. Performance rating scales, average merit increases and productivity will vary by a number of factors. Also, you may need to work closely with your finance department for a detailed understanding of how your business measures productivity by employee.
In summary, by using the correlation tool you can show which variables or factors have a relationship with each other and begin the path to building deeper predictive analytics.
You completed 0% of this lesson
You completed 0% of this course
Lessons Not Completed:
Christina A. Danforth, SHRM-SCP & SPHR, launched HR Jetpack in 2016 to support the development and professional growth of her fellow HR colleagues. She started her HR career in 2002. After obtaining her Master’s in Business Administration degree, Christina joined United Technologies Corporation. She moved across the United States...