Analytics, predictive analytics, diagnostics, data, big data and data mining are all terms used frequently in corporate training. But what do they mean, and what do training professionals need to know? Here are some definitions and tips for success.
Data, Analytics and their Uses
In training, analytics is the process of measuring an individual, system or organization’s performance. Training diagnostics is the process of examining and evaluating training and organizational performance through assessments, analysis and data collection. Big data refers to large, complex data sets that are difficult to analyze using traditional methods but that can reveal important patterns and relationships that inform decision-making.
Data mining and predictive analytics “are a collection of mathematical and computing techniques that can reveal new insights in data,” according to an email from Jeff Deal and Gerhard Pilcher, authors of “Mining Your Own Business: A Primer for Executives on Understanding and Employing Data Mining and Predictive Analytics.”
Data mining organizes data into patterns and relationships, and predictive analytics uses the data to make predictions about the future. Thanks to technology, artificial intelligence (AI) is increasingly used in this process through machine learning, which automates the process of analyzing data and making predictions. The people working with the computer, then, ensure the right questions are asked and, according to Peter Clark, co-founder of Qlearsite, use their experience to interpret what the AI tells them: “The combination of a human’s ability to ask good, relevant questions and an intelligent machine capable of searching within big data for a statistical answer is extremely powerful.”
Deal and Pilcher say that analytics can recommend new training and measure its effectiveness, suggest changes to existing training to improve outcomes, and measure the relationships “between training and employee retention” and “between training and employee satisfaction.” Data science, Clark summarizes, “will prove the ‘return on investment’ of learning … in short, [ensuring] the learning and development function delivers more value to the organization.”
James Densmore, director of data science at Degreed, elaborates, writing in an email that “Data Science is key to making the learning experience more personalized and more social.” Using “search and recommendation algorithms,” employees can learn the skills they need using the modality or modalities they prefer. Data can also help connect learners to their peers to learn from and collaborate with each other.
To reap the benefits of data and analytics in training, here are some tips.
1. Use relevant data.
“Just because it’s interesting doesn’t mean it’s valuable,” Densmore says of the data available to training professionals. Understand your business needs, and then define actionable metrics that will enable you to develop training to meet those needs. “Usually,” say Deal and Pilcher, “an effective baseline measure naturally emerges that becomes an effective tool for measuring ROI.”
Qlearsite uses natural language processing to convert written text into “analyzable scores on themes and sentiment.” Up to 80 percent of a business’ “people data” consists of communications, survey responses, performance reviews, assessments and other written text, according to Clark. It’s important to capture that information in a usable way.
2. Leverage technology wisely.
Algorithms “learn” from the data that you provide them with, and “algorithms are only as good as the data you train them with,” Densmore says. When using machine learning or other technology-enabled analytic techniques, evaluate your data and “tailor it to specific domains and use cases.”
3. Curate content thoughtfully.
There’s a large amount of content on the internet that learners can use effectively to improve their performance. Training professionals can help ensure that they access the right content at the right time using content curation. Densmore cautions that L&D organizations should classify content into interrelated topics rather than using a hierarchy, which doesn’t take the interconnectedness of topics into account.
Ask learners what they want to learn, determine what skills they need to learn based on their role, “look at the content they’ve been consuming” and then use all that data in your machine learning system to recommend relevant content.
4. Don’t oversimplify.
Clark says that “simple correlation of two metrics can create false signals.” For example, participants in one training program may have higher scores than participants in another, but does that mean the first program is better? It might be; on the other hand, the people taking that program may be more skilled than participants in the other program. Statistical factoring can alleviate this problem, and automation can help L&D professionals without a mathematics background complete that factoring more easily.
5. Communicate results.
Make sure the conclusions drawn from data are communicated to L&D managers. They need to know “what content is making the biggest difference, gaps in learning vs. industry trends and emerging technologies,” according to Densmore. That way, they can plan programs and resources strategically.
Data and analytics are powerful tools in corporate training, but only if they’re used strategically. Use these tips, and make your data work for you and your employees.