How AI is Affecting Modern Education – Will Your Child’s Future Teacher Be a Robot?

roman-mager-59976.jpg

The generation starting college this fall doesn’t remember a time before computers and the Internet. Their children may not remember a time before Artificial Intelligence.

Raised on technology, Millennials were the first generation to rely (at least partially) on the Internet to get their homework or college coursework completed. As technology continues to permeate the classroom, it’s only a matter of time before Artificial Intelligence becomes a key component in education. The opportunities presented by this development will be a decided departure in the way students learn and teachers teach...a modern twist on the timeless values of teaching.

Small Scale Disruption: Innovations in K12

If you’re familiar with Alt School, you’ll know that the future of modern education isn’t on the horizon – it’s already here. Alt School was developed by a former Google employee, Max Ventilla, who decided that algorithms driven by real-time data could lead to hyper-personalized learning for school children. This, in turn, would facilitate students to progress quickly through lessons on subjects they excelled at, and more slowly for those they found more challenging. Students are supervised by teachers, and their curricula are driven by the data fed into computer systems.  Alt School is just the tip of the iceberg in the move toward a more modern style of education.  Like many revolutionary approaches, it’s a small prototype and in its current form too expensive an approach for many school systems to implement.  

Disruptive Innovation at Scale: In Higher Ed

As Massive Online Open Courses (MOOCs) emerged in 2012 they were praised for allowing accessibility for those who couldn’t afford a traditional education, or those in remote areas who could not be physically present to attend classes. A scalable method of distributing information online, MOOCs are merely the start of the technological revolution in education.  

As they mature, these online approaches have met some resistance from critics of their comparative quality.  But, MOOCS and online courses fit the classic definition of disruptive innovation. Clayton Christensen of HBR says that disruptive innovations “are initially considered inferior by most of an incumbent’s customers. Typically, customers are not willing to switch to the new offering merely because it is less expensive. Instead, they wait until its quality rises enough to satisfy them. Once that’s happened, they adopt the new product and happily accept its lower price.” The key to increasing quality without blowing up the MOOC cost model?  High quality, scalable assessments that are driven by performance data.

And the opportunity that the explosion of online courses at traditional universities brings?  MOOCs serve as a precursor to AI and authentic assessments, by paving the way for students, educators and employers as they adjust to learning online in an environment where a massive amount of digital activity can be captured and analyzed.

Screen Shot 2017-09-08 at 11.09.03 AM.png

The Artificial Intelligence (AI) methods that Authess uses to measure complex problem-solving, are well-positioned to take a central role in higher education in the 21st century as MOOCs and online courses continue to take market share. The key to scaling these solutions while increasing quality is to make them affordable, efficient, and reliable. As the path of disruptive innovation continues, AI-driven personalized learning will be a true game-changer in online higher education.

Authentic Assessments and AI: Much-Needed Change

Benjamin Bloom’s “The 2-Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring” identified that 1-to-1 tutoring works best. MOOCs have delivered on the promise of reducing the cost of delivering instructional material, but they fall far short when trying to provide personalized learning and 1-to-1 tutoring experiences. The Holy Grail of online education is not just low-cost delivery, but truly personalized experiences with meaningful feedback and assessment capable of measuring critical thinking and problem-solving skills at scale. New developments in AI are opening the door to that kind of disruptive opportunity. Students can finally receive specific, actionable, and relevant feedback without a personal instructor sitting next to them every step of the way.

Data-driven technology takes things one step further by analyzing data from individual student performances and aggregating it, revealing trends and offering comparisons and benchmarks. This closed loop method benefits students by focusing their learning at a pace they’re comfortable with. It’s a dramatically more efficient system, too. Neither the student nor the educator waste time on what they’re already good at, but focus on where they are weak and need to up-skill.

Interacting with the course material and knowing when answers are correct (or not), is key to acquiring knowledge. It has been proven that learning is easier and better when students employ scenario-based approaches. Authentic assessments allow students to apply their knowledge to real-life scenarios. This not only reveals how well they have retained new knowledge, but also better prepares them for the workplace as they develop the skills to think critically and apply their knowledge in the real world.

Deploying these real-life scenarios at scale hasn’t historically been feasible, so we’ve fallen back on multiple choice and contrived short answer questions that only measure knowledge. These tests aren’t good predictors of students’ ability or true mastery of the concepts they are learning – just because students know factual material does not mean they know how to apply it to solving a problem. Employers meanwhile, care what you can do, not just what you know. There’s no doubt that authentic scenarios are a more effective means to develop problem solving and critical thinking skills, but these complex assessments are hard to administer and even harder to evaluate. AI is the disruptive opportunity that may finally be able to complete the circle and provide authentic assessment at scale.

Will Robots Replace Teachers?

One of the biggest arguments against the use of AI in education is that teachers will no longer be required. Advocates for AI disagree. Often, teachers’ capacities are overstretched due to the number of students they need to teach. They also seek ways to impart knowledge with as many of their students as possible, usually employing the traditional lecture-style approach—the least effective way to learn, according to Bloom’s research. But the worst part for teachers is knowing that, inevitably, one size lesson doesn’t fit all, and some students will fall behind.

The undeniable truth is that teachers can not personally grade the work of 500 students on their own. So, for larger classes or online programs, many have already replaced themselves in part by ceding assessment to auto-gradable multiple-choice tests. These tests have been shown to be poor measures of anything beyond factual recall. What Authess aims to do with AI and analytics is bring back performance-based assessments—the results of which are more meaningful and more representative of competency and mastery. AI doesn’t replace teachers, it augments them and provides them a powerful and effective tool to support assessment and feedback to enhance learning.

Shimon Shocken, founding dean of the Efi Arazi School of Computer Science at IDC Herzliya and advocate of MOOCs and education technology, believes “Teachers should be empowered, not replaced”. In fact, those who see the benefits of AI, authentic assessments, and education technology, argue that technological advancements in the classroom will in fact, allow teachers to excel at teaching. How?

For starters, AI solutions can provide educators with real-life information on who is performing above or below standards, allowing them to focus more time on coaching students who need it most, on the specific areas where they need it most. Thus, supporting Bloom’s point on 1:1 learning as being the most effective. Teacher interactions with students become much more relevant and productive to those students. One of the great potential benefits of AI and education technology is that they make it possible for teachers to spend more time with students on a meaningful, individual level, as the time grading coursework and lecturing is greatly reduced. If AI replaces anything, it’s the arduous and tedious act of grading. In its place, AI can provide teachers with higher level reports that focus on where the student missed the mark and where the opportunity for student development is most needed.

Identifying common problem areas for students can be accomplished efficiently by machines, but guiding how that weak area can be improved, of course, requires a human element. That’s when the teacher gets the opportunity to formulate a new approach to teaching that particular pain point.

Leveraging technology in this way also allows teachers to “flip the classroom”: students learn their lesson independently and then come together, with the teacher’s help and supervision, to apply that knowledge while interacting with each other. Individually learning lessons and then working collaboratively in a group is an effective way for students to apply the course content, and it also stimulates the development of power skills such as teamwork, critical thinking, communication, and problem-solving. All of these skills are commonly cited by employers as lacking in recent graduates. There is no age limit for this methodology either: it can just as easily be applied in a K-12 environment, all the way to adult learning.

AI and Education

Artificial Intelligence is already here, and whether we like it or not, students are living in a very different world now. Children born today never may never even drive a car.  The children of this year’s college freshmen will have a very different education experience to their parents and almost unrecognizable from their grandparents, but will share many of the same timeless values. AI-based solutions for education have the capacity to give these future minds the exact level of stimulation they need; the tailored attention from teachers they deserve; and the skills to succeed in their careers.

Why New Hires Don’t Stick

Systemic challenges in human resources and hiring processes are a key explanation for why employee turnover is so high. What can be done to address these issues?

Finding the right candidate to hire is a tough enough challenge, but having that employee flame out, is even more frustrating -- and costly.  The rate this happens though, is depressing -- the probability of a new employee failing after 18 months is close to 50 percent. The financial cost to companies should be an eye-opener, but other factors such as continued productivity, company reputation, camaraderie, and employee morale should not be neglected either.

Systemic Challenges

The reasons why new employees don’t always stay in their jobs vary greatly. Some are role-specific (working for a poor manager or accepting a lower salary than really desired) and must be addressed on a micro level. However, there are certain trends which are systemic to the recruiting and human resources sector, and which can’t be ignored by companies who aim to grow their business.

For starters, only recently have studies been conducted to assess the real cost of high employee turnover. Senior management is (naturally) only going to enforce changes in an inefficient process if it perceives a significant financial loss to the company. Because most companies don’t realize just how much they’re losing from this churn, they don’t feel the need for change.

In large part, recruitment is terribly inefficient because it lacks scientific or data-based assessments — companies rely too heavily on intuition and unstructured interviews, which are often flawed. A great comparison was made in The Guardian newspaper on how scouting for sports relies heavily on statistics and recommendations but how scouting for employees does not. In fact, studies have shown that interviews can do more harm than good.  Judging a candidate’s capabilities largely on interviews instead of on quantifiable data (academic merit, references, etc.) is more likely to lead to poor candidate selection.

Another systemic challenge is that hiring failures often aren’t tracked or quantified internally, which means that it’s impossible to know – and therefore address – the root cause of these failures. Learning where improvements can be made will lead to better candidates being hired, who are likelier to stay with the company longer.

Hiring processes (from applications process to interviews) and employers’ focus on hard skills (knowledge on a particular subject necessary to get the job done, such as proficiency in a particular software program) are unable to evaluate power skills or soft skills. It makes sense to guarantee that qualified candidates possess those hard skills. But in reality, the biggest complaint employers have — namely in regards to recent graduates — is that they are deficient in power skills such as critical thinking, communication, and teamwork.  These foundational skills are integral to success in any role, particularly in today’s workplace.

How to get new hires to stick

What if there were a way to make hiring more data-driven, and therefore more likely to select the most suitable candidates? Authess uses authentic assessment software to evaluate how candidates perform in real-life scenarios compared to benchmarks set by current employees in that company and role. These authentic assessments have two-fold value. First, they represent a clear indicator of a candidate’s approach to actual on-the-job scenarios, as compared to their potential future peers.  Second, these types of tests have shown to reveal more intangible skills like empathy, and power skills such as critical thinking, which are crucial to any role but which only become evident during this type of real-life scenario. Ensuring candidates have the right hard skills and power skills is one challenge these assessments overcome. Meanwhile traditional personality tests, in use by many talent acquisition professionals, have been shown to be poor predictors of job performance.  

Now, imagine a candidate has a degree in a relevant subject for a role, they interview well, and get the job. But without an understanding of their actual skills, there’s no way to know for certain whether they understand how to apply their knowledge in real-life. This is incredibly valuable knowledge for employers to have prior to making an offer. Undergoing an authentic assessment as part of the hiring process also exposes to candidates the type of functions they would be expected to perform, which means they’re less likely to have a false understanding of the job (another cause for leaving a new role).

If hiring processes--both filling the hiring pipeline with qualified candidates and narrowing down to the best candidates--were more scientific and data-driven, there would be a great deal more accountability (and trackability) from a cost perspective for the company. Additionally, because employers are provided with full reports on performance, they can identify positive or negative trends to address internally, if appropriate.

Looking ahead

Authentic assessments are currently being utilized by an array of sectors, due to the positive results they produce — especially in revealing skills that are nearly impossible to identify during the application process. As companies start to appreciate the value of recruiting the right candidates the first time around, more of them will start looking to data-backed tools like Authess to evaluate and redesign their human resources processes. Doing so will increase company profitability, employee satisfaction, and employee retention.

 

How Credo Education is Using Authentic Assessments to Measure Summer Interns’ Critical Thinking and Problem-Solving Skills

Discussing the theoretical value of authentic assessments is one thing; proving its efficacy in real-world scenarios is another. And that’s precisely what Credo Education, in partnership with Authess, has started to do.

Credo Education is a small, Boston-based company with big aspirations: to change the way higher education institutions prepare students for the workplace. They do so by selling foundational skills instruction and assessment solutions to those institutions. This summer, Credo will pilot a study using Authess’ authentic assessments to track and evaluate the skills development of 45 interns from various companies. These companies belong to sectors including educational technology, data and technology, digital media, and accounting and consulting. The real value of this pilot study will be revealed when foundational skills teaching and assessment is proven to be relevant across a wide range of sectors.

The Pilot Study

The process is straightforward. At the beginning of the summer, interns and their managers attend a 1-hour-long orientation. The interns then take a foundational skills-test on the Authess platform and their performance is assessed on a baseline (established by other employees) before their internships begin. The Authess assessment evaluates real-world problem-solving skills, critical thinking, reading comprehension, communication, and organization.  Over the course of approximately 8 weeks, interns go through the foundational skills program for an hour a week in addition to their internship tasks. At the end of the internship, the intern is assessed once again on the Authess platform to measure progress relative to their initial assessment, and a survey by both interns and managers is carried out. Credo then delivers a full report to employers.

Evaluating Foundational Skills

What are foundational skills and why do they matter? According to the National Foundation Skills Strategy Project, they are “employability skills, such as collaboration, problem solving, self-management, learning and information and communication technology (ICT) skills required for participation in modern workplaces and contemporary life.” Other examples of foundational skills include the ability to work in a team, to communicate professionally (both verbally and in writing), and to critically evaluate evidence. Time and time again, employers state how important these skills are, but for which many colleges fail to prepare their students. Most importantly for employees and employers, foundational skills are portable: they can be transferred from company to company and role to role. Employees with these skills are much more desirable to employers than those with purely hard skills (the skills specifically required to do a particular job, like knowing how to write CSS code, or speaking French). And just like hard skills, foundational skills can be learned.

Interns taking part in the study will learn these foundational skills throughout their 8-week internship and incrementally start applying them in the workplace. Interns who integrate the hard skills they learn in classes with the foundational skills they learn in the Credo coursework will have the best chance of success in real-world scenarios.

Benefits to Employers

The employers taking part in the study have much to gain from these authentic assessments. The obvious benefit is that their interns develop foundational skills much faster. A whopping 93% of employers surveyed by Hart Research Associates agree that “a candidate’s demonstrated capacity to think critically, communicate clearly, and solve complex problems is more important than their undergraduate major.” But in addition to having more qualified interns, employers receive quantitative reports from Credo on the interns’ performance before deciding which interns they plan to offer full-time positions to. Because the students are responsible for their own foundational skills learning throughout their internships, employers are not required to devote time and effort to teaching them these valuable skills.

Benefits for Interns

Naturally, the interns taking part in the study will also benefit greatly from the experience. Interns will finish their internships significantly more career-ready than they were two short months before – and they do not have to pay for this additional experience. Interns will receive a progress report on different areas of growth and be able to compare their performance with that of other participants in other companies. The certificate of completion they receive will serve as an added achievement on their resumes.

Authentic Assessments

One of the greatest benefits of Authess’ authentic assessments is that they serve to identify and develop crucial foundational skills, which are transferable to any undertaking. It is precisely because they are transferable that results are expected to remain repeatable, regardless of industry sector. This pilot study will be completed by the end of summer 2017 with published results to follow. It will be interesting to evaluate the impact these authentic assessments will have on interns’ foundational skills development, and learn more about the value employers stand to gain.     

 

If you are interested in learning more about the Credo pilot study, please contact askus@credoeducation.com. For more information about Authess, please visit their website or contact info@authess.com

Reducing Employee Turnover: Why the Art of Recruiting Candidates Should Be a Science

Losing an employee is financially and logistically costly. High employee turnover places added pressure on recruiters to find the “ideal candidate” who will remain and prosper with the company. Herein lies the HR challenge: recruiting management employees and reducing turnover.

The Challenge Presented by the Status Quo

The problem of turnover is linked to societal conventions: The usual methodology for recruiting involves collecting resumes and cover letters from candidates, screening (by a human or a computer), short-listing for interview, rounds of interviews and final candidate selection. Of course, it would be unthinkable to hire a candidate without interviewing them. But according to the research reported in a recent NY Times article, interviews are flawed and can even hinder the hiring process.

Candidates rehearse answers to anticipated questions, preparing responses that they know the employer wants to hear. Candidates also lie about their experience (because they can get away with it) and knowledge. Richard Branson is famously quoted for saying, “If somebody offers you an amazing opportunity but you are not sure you can do it, say yes – then learn how to do it later!” The problem is, future employers don’t want candidates who say they know how to do something when in fact they don’t.

Naturally, this qualification gap is felt more strongly in certain roles than in others. A graphic designer, for example, may have to produce a sample piece as part of their interview process, but it’s rare that a management employee, for example, would be tested on their skills. But the question is: why aren’t they? Wouldn’t this address so many of the competency and knowledge – and to some degree – cultural mismatches that are a large contributing factor to high employee turnover?

The Opportunity to Change the Future

Authentic assessments, where candidates are tested on real-life scenarios specific to a particular role, may be the future of recruitment. Much like testing cars on the road before choosing the one you buy, testing candidates before offering them a job would ensure greater compatibility between the role and employee.

Those who have used authentic assessments for training and evaluations have been impressed by the results. The level of knowledge and experience in subject matter quickly becomes evident. But, interestingly, these assessments also are also capable of detecting nuances and “power skills” that a particular employer might be looking for in a candidate. For example, stricter/more lenient decision-making, or prioritizing one particular detail over another can reveal skills like judgement and critical thinking. Especially in managerial positions, these subtle differences can differentiate between a competent candidate and a great one. Differentiating between good candidates and great ones using only interviews and a resume involves an undefined level of risk: people (yes, even professional recruiters!) are often misled or influenced by factors which are irrelevant to the job and, surely enough, mistakes are made. Some people can certainly “talk the talk”, but when it comes time to “walk the walk”, they fall behind.

The Power of Power Skills

Research carried out by LinkedIn suggests that close to 60% of companies struggle to find people with the skills they need. Power skills like critical thinking and problem solving (chief concerns of employers who hire millennials), communication, conflict resolution, teamwork, and adaptability are regarded by many employers as just as crucial to long-term success as job-specific knowledge or skills. On paper, a candidate may appear to have all the right knowledge, but an interview can’t tell you for sure whether they will be able to collaborate with team members, or work efficiently to get the job done. Essentially, will the candidate fit in to the company culture? The more confident, motivated and satisfied an employee feels in their role, the less likely they are to leave. Authentic assessments provide insight into power skills like critical thinking and adaptability: how well a candidate can assess or adapt to a situation, how they choose to use available information to do so, and how that compares to benchmarks of expertise in the field or even other candidates.

So, while employers should seek employees with the necessary technical knowledge to do a job, they shouldn’t underestimate the importance of power skills when it comes to big-picture goals like employee retention and satisfaction. The particular duties within any job, especially in management positions, are bound to change and evolve; power skills are what enable the necessary flexibility and adaptability for long-term success.

If training and evaluating management employees were treated less like an art (learning about someone through a back-and-forth Q&A session) and more like a science (backed by algorithms and AI), the world would undoubtedly be filled with more competent and satisfied employees, fewer financial losses for companies, and more assured recruiters.