Dr. Max Blumberg

Why senior management don't often rely on empirical research

2011-08-26T09:26:00.001+01:00

In a post below, I asked how much value is added by people-related activity in organizations using this model:

http://img.photobucket.com/albums/v480/maxblumberg/PeopleValue.png

The model can be used to assess how the extent to which people-related activity adds to Total Shareholder Return (TSR). TSR reflects total equity growth and is the ultimate outcome measure chased by senior management in most `commercial organizations*.

I suggest that we have little evidence about the value added by People in many organisations. By People, I mean employees (payroll, workforces, human capital, talent) and people-related investments (cradle-to-grave: attraction, selection, onbboarding, development, succession planning, engagement, retention, exit). The size of People investments vary by industry, but as long ago as 2002, CFO Research Services reported that as a percentage of of revenue, it was 37% in TMT, 25% in Heavy Manufacturing, 45% in Pharmaceuticals, and 43% in Financial Services. I can only suspect that investments have increased since then. Thus a Financial Services company with $5bn turnover would be spending $2.25bn on People.

This is a lot of money and if the statistic is correct, presumably senior management believe that this investment is yielding a better TSR than investment in alternative investments. Again, as suggested in a previous post here, this diagram presents alternative possible investments that senior management could have made to maximise TSR:

http://img.photobucket.com/albums/v480/maxblumberg/Humancapital.jpg

Senior management must therefore decide which mix of investments in tangible assets, intellectual capital, and financial capital is likely maximize TSR (in other words, it is a portfolio management problem). On what do they base this decision? As scientists, we would suggest that the effective method is empirical research. But senior management are nervous of research for the following reasons:

o Robust valid research is often difficult – if not impossible – to achieve in the ‘real world’ because of the reasons cited above (essentially, there are simply too many variables, most of which cannot be controlled to yield robust valid results)

o Research is expensive

o Research takes time and by the time outcomes are known, investment opportunity window may have disappeared or a competitor may have taken the market sooner by making a (admittedly riskier) non-research based decision

o The research may reveal that the asset class (e.g. tangible assets, human capital, etc.) does not in fact yield the greatest returns.

For this reason, instead of relying solely on empirical research, senior management tend to rely on a blend of past experience, intuition - and perhaps some limited research. This outcome may be disappointing to those of us who favour empirically-based decision-making, but that’s the real-world for you.

I would therefore suggest that, with good cause, real-world management is probably less dependent on research for their weighty decisions than one might like to think.

*(It can be argued that Balanced ScoreCard and Corporate Social Responsibility should also be included as ultimate measures, but these are probably better viewed as mediators of people-related activity on TSR - perhaps a discussion for another day).

What counts as evidence?

2011-08-25T13:30:00.005+01:00

Ron Kennedy as ever asks great questions. Today he asked, what counts as evidence in the world of organisational psychology? My response:

Does this have the potential to be a religious war? Qualitative researchers are not always going to accept the positivist's evidence and vice versa. So prior assumptions about reality seem to play a role. For example: Can anything be objectively measured? Does the presence of the researcher affect the outcome of the investigation? Is the outcome a linear consequence of the cause or does reciprocity play a role? All of these affect the weight of evidence presented and should be acknowledged by the researcher.

From my perspective, what constitutes evidence depends on what is being investigated. Let's say a client wants to know which organisational and psychological attributes drive employee retention in their investigation. And say I am able to build a model whose predictors distinguish between stayers and leavers with 75% accuracy. On the surface, this is remarkable since the probability of achieving 75% accuracy by chance is just 2.81 x 10^-7 (binomial theorem) i.e. close to zero. But before celebrating my evidence, I must ask:

o Was this based on genuine groups of stayers or leavers or did I base it on the tenure of existing employees? If so, my evidence is weakened.

o Does this reflect the existing employee population? Can I generalise it to future populations accounting for the attitudinal shifts of Gen Y & Z etc? If not, evidence is weakened.

o Might my findings be 90% accurate for one business unit, but only 60% accurate for another? If I haven't controlled for this my evidence is weakened.

o Might external events in time affect retention (e.g. economy, M&A, etc.) have affected it? If I can't control for these, my evidence is weakened.

o And so on.

So evidence is only evidence after all factors which might affect it have been weighed and what on the surface appears to be evidence for a large effect size may dwindle down to evidence for not much at all.

On the other hand, if I need evidence for how employees feel about a newly proposed pension scheme and I am not trying to predict anything, I could probably interview a representative sample of sufficient size and come up with reasonable evidence one way or another.

Bottom line: Evidence resulting from predictive models should be treated with extreme caution and all limitations and assumptions noted. Evidence resulting from descriptive investigations is probably quite robust providing samples are representative and of sufficient size.

Original post:

http://www.linkedin.com/groupAnswers?viewQuestionAndAnswers=&discussionID=67058084&gid=78865&commentID=49666523&trk=view_disc

Superficial analysis and the wrong interventions

2011-08-16T08:45:00.001+01:00

Labour opposition leader Ed Milliband says the UK needs a proper, deep analysis into the real causes of the recent riots. But Prime Minister David Cameron says the UK knows the causes and all that is needed now is rapid action and intervention to prevent recurrence. Milliband responds to this by saying that this means Cameron's intervention will simply be a knee-jerk reaction based on a superficial analysis of the situation, and which will result in the wrong intervention and not cure the rioting.

As diagnostic analysts, we see this situation almost daily in complex organisations. We hear about their problems finding the right employees, and how they spend a fortune on development, engagement and retention, but without the results they were hoping for. But when we offer to perform a formal diagnostic to identify the root causes of the problem, there is inevitably a Cameron in the company who tells us the answer is obvious and then proceeds with an intervention. Nine times out of ten, we're back there within two years.

It is true that in non-complex organisations, the causes of problems are indeed usually easy to identify. But in large, global, multi-cultural situations, they are not.

Why do people fear diagnostic analysis? Some possible reasons:

1. They cost money, it is true. But they are usually cheaper than applying the wrong intervention.

2. Some people get systematic problem analysis; others don't. I notice this often based on the looks of awe that I get when explaining what to me seems like a simple systematic approach to, for example, identifying the causes of employee turnover. I often wonder why the client didn't think of it themselves, but the fact is that many people do not think this way and find it difficult. I suppose it's good because that creates opportunities for me.

Of course sometimes even formal diagnostics can get it wrong, but one thing is for sure: they get it wrong less often and cost much less than superficial guesses about the underlying causes.

Does human capital inevitably yield higher returns than other "assets"?

2011-08-05T09:28:00.002+01:00

Commercial organisations may wish to consider this model:

http://img.photobucket.com/albums/v480/maxblumberg/PeopleValue.png

It argues that people decisions ultimately underpin total shareholder returns (TSR). Specifically, the effect of people on TSR is mediated by Quality, Innovation, Producvity and Customers. (Few consultants actually go so far as to try and measure this - but they really should to justify their interventions).

But people really the biggest TSR driver in most organisations? Consider instead this asset-based model of the enterprise:

http://img.photobucket.com/albums/v480/maxblumberg/Humancapital.jpg

This model argues that the sum of tangible assets, intellectual capital (including human capital) and financial capital drive TSR. Can you swear for sure that the effect of human capital accounts for most of the variance in TSR? For example, in a very automated environment, it may be that machines account for more TSR than people. Or in a financial services environment, perhaps financial assets generate larger returns than people. For example, I remember the CEO of Barclays once telling to me (long ago in the 80s I grant you) that their financial asset base was so large that it made money in spite of employee interventions. I asked what he meant by this and he explained that if it was mainly when employees meddled with their huge financial assets that they lost money!

The most common argument for the power of human capital - the domain of the IO psychologist - is that ultimately, all decisions about enteprise assets - acquisition, maintenance, and disposal - are made by people. But can we acknowledge at least that sometimes, just sometimes, non-people assets can generate higher returns than the people who set them up?

Is it possible to be evidence-based and rigorously scientific in the real world?

2011-08-02T07:47:00.003+01:00

"Is it possible to be evidence-based and rigorously scientific in the real world?"

If evidence-based and rigorously scientific means applying scientific method and reporting any limitations to the client, then yes, it is definitely possible. Even in the academic world, one cannot always use that gold standard, the randomised controlled trial. In this case, one simply ensures that readers understand this by clearly noting it as a limitation of the research. Similarly when one encounters limitations in real world research, all we have to do is make our clients' aware that there may be some risks in adopting our recommendations (selection, development, re-org, etc.)

And I believe that limitations reporting is an area perhaps where perhaps real world I-O practitioners could do better. We don't always tell our clients that this psychometric was normed on a population different to yours, that the original results came from a prospective rather than a retrospective investigation, that we can't be 100% sure of the conclusion because correlation is not causation, and so on. And maybe with good reason: because the clients may stopping buying if they knew this!

Bottom line: Evidence-based practice is an ideal seldom achieved in the academic world, and even more difficult to achieve in real world (because conditions are even less controllable). But the role of scientist-practitioners is to implement it to the best of their ability while making sponsors aware of any limitations.

The difference between metrics and analytics

2011-07-29T09:56:00.001+01:00

Metrics
======

Metrics are simply measurements, for example "Our average engagement level is 80%", "Our average employee attrition is 30%", "Our average performance level is 60%" and so on.

The problem with metrics is that they can't tell you whether the number is "good" or "bad" for the organisation. For example, if someone says attrition 30%, that sounds bad. However, if those 30% are mainly low performers, that is probably a good thing. Another example: if average engagement is 80%, is that good or bad? Is it better than engagement of 70% and worse than 90%? You simply can't say until you know for sure whether increased engagement leads to increased performance. For example at Friends Provident, we found that performance peaked when engagement was at 70%, but declined as it increased beyond this level. This may be because excessive engagement leads to exhaustion or because managers are willing to sacrifice high performance in the name of high engagement.

Either way, metrics in themselves don't tell you what actions are needed to improve performance.

Analytics
=======

In HR terms, one could say that human capital analytics examine the effect of HR metrics on performance. In more general terms, analytics look for patterns of similarity between metrics e.g. do competency levels and retention levels increase at similar rates? Do competency levels and performance levels work together?

If analytics show that two metrics *do not* work together, it is pointless for the organisation to invest in one metrics in order to change the other. For example, if analytics shows no relationship between salary and retention rates, then it would be pointless investing in payroll changes. And it means the organisation must search for metrics that *do* have a relationship with retention. Yet it is amazing how many organisations invest millions in programmes to change some metric (engagement, retention, competencies) without first ensuring whether that metric really does have an impact on performance. This is usually money poured down the drain and are the expensive programmes one hears about that failed.

Bottom-line: Metrics are just measurements; analytics show the relationships between metrics and suggest what organisations can do to improve performance.

Organisational psychology in the academic versus the real world

2011-07-29T09:51:00.000+01:00

A number of people suggest there is no difference between academic and organistional research and that conclusions from academia can be easily transported to organisations. I argue that this is untrue for the following reasons:

* The academic world optimises the experimental environment to achieve MAXMINCON: MAXimise variance due to the independent variable(s), MINimise error variance, CONtrol nuisance variables and extraneous variance. But the real world is not some controllable experiment where one can reconfigure the client's organisation to achieve MAXMINCON. Circumstances change half-way through an exercise (economy, new CEO, etc. - all these impact measurements and interventions).

* The sample sizes available in the real world are often too small to deliver the required power to draw the heady conclusions available in academic research (where we keep on building the sample size until it is large enough)

* Error variances are seldom normally distributed in the real world and are related. Most techniques learned in academic programmes are based on a General Linear Model which when applied in the real world lead to inflated alpha and beta errors - or in English, are often not valid. Different analytical techniques are required in the real world in order to deliver 'evidence'.

* Conclusions gleaned from academic research are valid under the same controlled environment; making claims that they can be generalised to real world organisational environments is irresponsible. For example, if one finds in academic research that a given competency framework results in higher performance in an experimental group relative to a control group, one cannot simply sell this framework to every organisation claiming it will work as well. Workforces, cultures, regions and buiness units differ.

* In the academic world, even small effect sizes are acceptable as long as the result is significant (say p < .05). Say you find a correlation of r = .20 (p <.01) between engagement and performance. In the academic world, this is a publishable win even though only 4% of the variance in performance is explained by engagement. In the real world, are you really going to suggest to head of HR they should invest in an expensive engagement programme - knowing that engagement accounts for only 4% of performance? I often see this being done in the real world and it is at worst unethical and at best shows non-understanding of the difference between significance and effect size.

Bottom line: One needs to be circumspect about claims about evidence-based interventions in the real world.

Recruitment agencies: Client interests at heart? Think again!

2011-03-29T09:41:00.000+01:00

They say that when jackals start worrying about the welfare of sheep, you know there's an election coming. So I guess I was naive to ask a recruitment agent whether he was interested in a joint venture selling my system for accurately identifying the right employee for the job.

The system is simple enough (except for a bit of statistical analysis): you psychometrically test a sample of employees from the role in question. Then you work out which personality traits, competencies and values separate the high from the low performers. New applicants applying for the role then do the same test and you simply select those who look more like your high performers.

This approach gets it right about 70% - 85% of the time which is a hell of a lot better than flipping a coin which is what businesses are effectively doing by using off-the-shelf psychometrics and untrained interviewers. But the proof of the pudding is that clients typically report 15% - 30 performance improvements. Whether this is our system or simply because it makes people pay more attention to new personnel is an interesting debate, but the point is - it gets results!

But perhaps asking a recruiter to partner to sell it was naive. As he pointed out: "Max, your system might reject applicants that I put forward".

"Well, doh, yes" I said, "Isn't that the idea? We reject the candidates who look like low performers and progress the ones who look like high performers".

"But that's bad business for me because then I need to go back and try to find them a high potential candidate. That's more work for me".

"So are you saying you're happy to take the money from the client even if you aren't providing them with high performers?"

And that was the end of our conversation.

The bottom line is that good science is effective, but it won't win the support of intermediaries who make money whether or not they perform.

Sergey Brin

2008-12-07T09:07:00.000Z

Sergey Brin - founder of Google - learned that he could be significantly predisposed to Parkinson's Disease. When asked whether ignorance might have been better, his response is that knowledge is better than uncertainty since he can now make adjustments to his life to reduce the risk and fund further research. I suspect it is this kind of thinking that puts Google where it is today; it's all about Knowledge.

Yet, many organisations prefer to either live in ignorance about the risks in their organisations or rely on subjective gut-feel rather than objective measurement to assess it.

Of course the advantage of measuring risk (operational or strategic) is that the act of measurement itself often points towards the very thing that needs to be adjusted to reduce it.

Yet many businesspeople are frightened of measurement. Why? Because it's benefits are uncertain? Because they are afraid of their inadequacies being exposed? Because they're afraid of looking foolish because they don't understand the measurement techniques?

The interesting thing is that both Brin and Gates are inherently measurers, and in both cases, it's paid when they bother to do it.

HR Silos

2008-04-10T09:28:00.000+01:00

The term silo has a well-known negative connotation suggesting uncordinated activity to the detriment of the organization. HR people are particularly keen on silo-busting because it inhibits implementation of best HR practices.

But HR functions can also fall pray to silo-creation if the various specialists functions do not work closely together and lead to unintended consequences. For example in many organizations, employees are rewarded for their individual effort and receive some kind of bonus if the organization as a whole performs well.

However, an employees' locus of control usually extends to their own work and to that of their team. Unless they are a board-member or the CEO, they can seldom influence the organization as a whole and therefore company-wide bonuses (like share-schemes) are flawed from the start.

A more pragmatic approach is therefore for reward and performance-appraisal specialists to de-silo and work together to create rewards that are a function of individual and immediate team performance.

Such systemic effects can occur between all HR functions (employee development, relations, succession planning, and so on) and will be addressed in future posts.

Return on People versus Return on Capital

2008-04-10T09:20:00.000+01:00

Shareholders invest in companies. Companies invest in assets. Shareholders receive the returns generated by these assets.

These assets can be viewed as the well-known "factors of production" required to create wealth. The basis for wealth until the industrial revolution was land. The revolution changed the focus to capital goods and technology which served us well until about the 1980s.

We are now climbing steadily into the knowledge economy and the focus is shifting rapidly towards people as the creators of organizational wealth. It is also argued that talented people are becoming relatively more scarce than capital or technology. (More on this in future posts).

So whereas before, it made sense to consider "return on capital employed" where that capital was land, financial investments, or capital goods, there is an growing need to measure "return on people employed". As Lowell Bryan of McKinsey puts it, profit per employee.

But whereas it is relatively easy to assign profitability to individual capital goods (machinery), funds, and technology, it is not as easy to determine the profitability generated by each individual in an organization, critical though this knowledge is. Without it, we might be under-investing or over-investing in people. Human capital measurement is therefore critical for maximizing earnings.

Currently, the best way of measuring return on people investments is to do it on a profit centre by profit centre basis. That is, determine how much was invested in people within a given profit centres, and then try to determine how of the profit was due to investments in people as opposed to investments in other intangible assets, or in capital and technology.

In the coming weeks, this blog will examine approaches for teasing apart how much of the profit can be attributed to people as opposed to other assets.

As an aside, investments in people are not really classed as investments at all. They are expensed, probably because organizations cannot "own" people in the same way they own other assets. Organizations rent people and the profits generated by individual is the return on rent paid. Furthermore, it is argued that the disparity between a company's market value and its book value is due to its intangible assets, one of which are the people that it "rents".

More on this later.

How to manage Generation Y employees

2008-04-06T14:06:00.001+01:00

A fascinating article on The Manager talks about how to manage salespeople born around 1986 - 2000. In fact, the article content is useful advice for any older person who must manage younger people.

As with most articles about Generation Y, the key message (for me at any rate) is about their loyalty and how much harder it will be to retain them. What are the secrets to retaining them? For one, they are "information" generation to whom the Internet is second nature. Therefore, one way to manage them would be to allow them to reach their own conclusions by letting them seek the information they need. Older team members, on the other hand, are more used to being "told" what to do. In fact, this is another area where the old could profitably learn from the young.

Enjoy the article.