Sunday, November 28, 2010

Statistics for HR...Correlation, Regression, OH MY!






In most of my teaching and consulting in the area of Strategic HR, I discuss the importance of HR having or obtaining analytical skills as I truly believe we are moving to an HR model that will look much different in the next few years. (HR 2.0)

I always get the question, "What skills do I need and where do I get those?"

First, I will talk about "what skills." I think basic statistics is necessary in HR to analyze all the diverse data sets in organizations as the data relates to human capital. Here are the must haves:

1) Correlations-are used to understand how data sets are related. In other words, if variable A changes does variable Y change? There are about a million ways this can be used in HR alone. If engagement goes up, does turnover go down? (negative correlation). But in a broader sense, you can analyze the relationship between employee behaviors and customer behaviors. If employees are knowledgeable regarding our products, do sales go up? Correlations can be calculated very simply in Excel or SPSS.

2) Regressions-Regression analysis is widely used for prediction and forecasting. Regression is also used to understand which among the independent variables are related to the dependent variable, and to explore the forms of these relationships. In restricted circumstances, regression analysis can be used to infer causal relationships between the independent and dependent variables. So, correlation tells you if a relationship exists, regression tells you which variables have the most impact on the dependent variable. So, in an example from our work, we look at engagement data to find out what variables have the MOST impact on employee engagement for a particular company. As these variables can be different from company to company it is important to know what drives engagement so you can keep doing the right things. Regressions can also be calculated in Excel and SPSS.

Where can you get these skills?

I am living proof, it can be done. My analytical journey started in 1995 when I reported to the CFO. My days were over when I could embark on the HR program of the week without a business case for why I wanted to create this program and WHAT the expected impact would be.

Fast forward, to stating our own business and working with our client's multiple data sets. We had some very interesting questions that needed statistics to answer properly. So, I enrolled in statistics classes at Georgia State University, where I made the highest grade in a class comprised of math majors. (my business partner was 2nd!) I don't tell you this to brag (well maybe) but to say, we had a distinct advantage over the math majors. We knew how to talk the language of the business and how to tell a story with data...they did not.

Now, back to HR 2.0 that will require analytical skills. It can be TAUGHT IF YOU WANT TO LEARN. But, they can be BOUGHT if you don't want to learn as well.

Bottom line....learn them or buy them....you needed these skills yesterday.

10 comments:

Chris Young said...

Nice post Cathy - You are dead on that HR pros need these skills yesterday!

I have included your post in my Rainmaker 'Fab Five' blog picks of the week (http://www.maximizepossibility.com/employee_retention/2010/11/the-rainmaker-fab-five-blog-picks-of-the-week.html) to give my readers a wake up call on the importance of statistic skills in human resources.

Be well!

- Chris Young

Brentano the Consultant said...

Good piece.

Words of encouragement: HR professionals should take advantage of more statistical training.

Now here's the challenge to everyone: proceed cautiously with your new education in this arena. Some believe they know enough now to 'drive the car'; however, they have merely been shown the gas pedal, may be blindfolded, have their hands tied, or are unfamiliar with the controls. I have witnessed quite a bit of misguided consulting upwards to leadership about what the data actually represents, the inferences drawn, and gross miscalculations of behaviors to outcomes. If you are in a position to consult leadership on programs, processes, or climate inside the organization, it is advisable to get very in-depth with advanced research methods or statistics offered at the behavioral sciences' graduate level (i.e. multiple regression, path analysis, structural equation modeling, multivariate analysis, quasi-experimental design, etc.). Or, leave it with an expert and at least be a partner 'in the know' when you're at the table. This comment is not intended to be intimidating to folks.

Cathy Missildine-Martin, SPHR said...

Chris, I appreciate your support, you always have such nice things to say and I really appreciate it. Have a great week!

Brentano:

I couldn't agree with you more. I have seen that in action many, many times. I love all those advanced methods you mentioned as we have had to use those depending on the project. BUT, you have to KNOW what the data says and you have to UNDERSTAND the context. I find that people run the advanced statistical test and then are unable to tell the data story in a way that is meaningful to the business. I do't think you were being intimidating at all...just honest!!!

Thanks for reading, I do appreciate it.

Cathy

Brentano the Consultant said...

Another benefit to the HR professional in understanding advanced statistics and research methodologies: it will be easier to take a critical eye on whether the vendor sitting across from you has a solid product, and help you detect unbridled embellishment on its features.

Cathy Missildine-Martin, SPHR said...

Another good reason..to keep people like me inline!!! Thanks Brentano for keeping the conversation going!!

paraman52 said...

Good write up of basic, must know, analytical skills.Well done! It proves that there is no excuse for not knowing some statistics and research methodology. The best way to prove a point - and sell a program - is to back up your conclusions and recommendations with hard facts and data.

Cathy Missildine-Martin, SPHR said...

Thanks paraman52 for reading! I believe decision in business are made more on data than gut feel post recession....so now more important than ever!

Ms. Michael Moon said...

I appreciated this post a lot. As a fellow HR professional with a background in statistics and network analysis I know deep down inside I have the skills that HR needs, but convincing them they need people like me has been an uphill battle. I even went for a PhD in hopes that would give me the extra credibility in the eyes of the business. Not so. In many of my roles, i have watched as we misinterpret survey data and draw incorrect conclusions and when I attempt to draw attention to the holes in the data collection or conclusions based on inappropriate sample sizes or lack of significance testing I am looked at as if I have three heads. Definitely an uphill battle.

What do you think it will take?

Ms. Michael Moon said...

I appreciated this post a lot. As a fellow HR professional with a background in statistics and network analysis I know deep down inside I have the skills that HR needs, but convincing them they need people like me has been an uphill battle. I even went for a PhD in hopes that would give me the extra credibility in the eyes of the business. Not so. In many of my roles, i have watched as we misinterpret survey data and draw incorrect conclusions and when I attempt to draw attention to the holes in the data collection or conclusions based on inappropriate sample sizes or lack of significance testing I am looked at as if I have three heads. Definitely an uphill battle.

Cathy Missildine-Martin, SPHR said...

Michael:

I see the tide turning on HR hiring statisticians! I have more and more conversations regarding that is a needed skill set within HR. I think HR has been a little "snobby" lately about requiring HR experience as well. I don't think that is necessary. I believe if you have good business sense and can ask good questions and know how to perform analysis then you can learn the HR stuff.

Send me your resume I have a client that is looking right now. cathymissildine@intellectual-capital.net

Cathy