Identifying Business Chemistry Types

There are two main ways to identify a person's Business Chemistry type'the low-tech way, which you can apply right now after reading this book, (or even glancing through it), and the high-tech way, which we use in our research studies and client engagements. We'll explain both here, starting with the low tech.

If you've read all the chapters of this book, you've learned everything you need to know about the Business Chemistry types and how to recognize them. At this point we bet you have your own type identified and those of the people around you. If for some reason you're reading this appendix first, just take a close look at the summary descriptions of the types in Chapter 1, or at the figure located directly after Chapter 6, and pretty quickly you'll likely know your own type and that of others. If you haven't already, you could also read Chapter 7, where we describe the process for developing a hunch, which can be applied to identifying your own type or someone else's. Speaking of developing a hunch about someone else's type, we've also created an online tool designed to help you do that. The tool guides you through 20 questions that narrow down someone's likely type.

While most of you will identify your type through the low-tech method, since our research is based on data collected with the high-tech method we want to provide you with some of the details about that here. Essentially, people use our online Business Chemistry tool to respond to each of the 68 assessment items by using a slider to indicate how accurately the item does or does not describe them. And then a bunch of math happens and they get a report telling them their type.

What kind of math you ask? (If you're reading this appendix you're likely one of the types that would ask that question.) We're happy to oblige; two main calculations happen.

First we normalize a person's ratings of each item to account for individual response biases. We do this because when taking a survey some people tend to respond to everything in a very extreme way while others tend to be a bit more moderate across the board. There's nothing wrong with that, but what we're trying to get at is how a person responds differently to one item versus another (not whether or not they respond in an extreme way to everything). Basically, we adjust a person's scores in a way that accounts for this (we use their average and standard deviation across all the items to do so, if you really must know).

Once that's done we go back to those Eigen values we told you about a few pages ago and use them to transform a person's normalized scores for each item into raw scores for each dimension. Those raw scores tell us whether a person is closer to the Pioneer or Guardian end of one dimension and whether they're closer to the Driver or Integrator end of the other. Now we're getting somewhere.

Finally we compare a person's raw dimension scores to our baseline sample to determine in which percentile they fall for each of the types. You can read more about the composition of our baseline sample below, but basically it's a sample of about 1,500 men and 1,500 women of varying organizational levels who work in the United States in one of more than 700 different companies representing a variety of industries.

Why do we do this? As we discussed in Chapter 7, Business Chemistry is relative. That is, there's no such thing as a Pioneer (or any of the other types) in a vacuum. A person is a Pioneer in relation to how Pioneer-like those around them are. The same goes for the other types. That means someone may be a Pioneer in comparison to the typical professional, but an Integrator compared to the typical leader. In a similar way, they may be considered tall in a room full of kindergarteners but short in a room full of basketball players, even though their height hasn't changed.

A person who completes the Business Chemistry assessment receives a percentile score for each of the four types in relation to this baseline, with the highest percentile representing their primary type, and the second highest their secondary type. They also receive percentile scores in relation to the Dreamer, Teamer, Commander, and Scientist subtypes.

Baseline Samples (Featured in Chapters 9 & 11)

Our Business (U.S.) Baseline sample is used in two primary ways. First, we use it as the comparison sample representing the typical professional when we calculate respondents' Business Chemistry scores and identify their types'the process we just described in the previous section. Second, we use this sample to explore introversion/extroversion (described in Chapter 9) and Business Chemistry differences by gender, generation, and organizational level (all described in Chapter 11).

The reason for using this sample for these analyses rather than some of our larger samples is that this one was specifically selected and stratified with particular attention paid to how variables may overlap. For example, we made sure that there were roughly equal numbers of men and women at all organizational levels so we didn't end up with a confounding effect of higher levels being male dominated. Our other samples aren't selected, rather we include everyone who responded to our questions. We often repeat demographic analyses with our larger samples to confirm the patterns we identify with this baseline sample, but we report results from this sample because it's cleaner.

Our Business (U.S.) Baseline is comprised of 2,958 U.S.-based professionals, 50% male and 50% female. We included professionals from a representative mix of 714 companies across industries, ensuring that no particular company made up more than 5% of the sample.

The margin of error for this sample is less than +/−2 percentage points at a 95% confidence level. In other words, we can be very confident (95% confident!) that were we to assess Business Chemistry across a large swath of the U.S. professional population, the proportions of the Business Chemistry types would be within a couple of percentage points of what we're seeing in this sample (which is a relatively even distribution across types).

One caveat here is since our sample was deliberately selected rather than chosen randomly, we might expect patterns to differ a bit more depending on the overall demographic make-up of the large swath we assessed. For example, while Millennials are now the largest generation in the U.S. workforce,¹ our Business (U.S.) Baseline sample includes fewer millennials because we intentionally selected for more professionals at the manager and leader levels. Our other samples all include higher proportions of Millennials.

Margins of error for men and women in this sample are +/− 3, and for organizational level and generation as follows:

Business (U.S.) Baseline Sample Demographics

In addition to the Business (U.S.) Baseline, which is our default baseline and offers respondents the chance to understand their Business Chemistry type relative to the typical professional, we have a second, optional baseline. Our Leadership (U.S.) Baseline allows respondents to compare themselves to the typical leader. Because remember, as we discussed in Chapter 7, Business Chemistry is relative.

Our Leadership (U.S.) Baseline is comprised of 1,000 U.S.-based leaders, 52% male and 48% female, serving at the director level and above. We selected leaders from a representative mix of 453 companies across industries, ensuring that no particular company made up more than 5% of the sample.

Leadership (U.S.) Sample Demographics

The margin of error for this sample is less than +/−4 percentage points at a 95% confidence level. We have not reported margins of error for the demographic categories because we do not use this sample for demographic comparisons, but only for calculating Business Chemistry scores in relation to the typical leader.

Stress Study Samples (Featured in Chapter 10)

Our stress study includes two samples of professionals working at varying organizational levels in the United States and elsewhere. Participants represent more than 1,300 organizations across various industries, including Deloitte, and 120 countries overall.

Stress Sample 1 is comprised of 23,597 professionals who, during the period of November 2014 to June 2015, completed the Business Chemistry assessment online and also answered questions about their current and general stress levels, and about how stressful they find 15 workplace events and situations to be.

Stress Study Sample 1 Demographics

Stress Sample 2 is comprised of 17,008 professionals who, during the period of June 2015 to October 2015, completed the Business Chemistry assessment online and also answered questions about how effective they are under stress and how often they use 12 different coping strategies.

Stress Study Sample 2 Demographics

The margin of error for both samples is less than +/−2 percentage points at a 95% confidence level, for all Business Chemistry types, including subtypes.

We do not have sufficient information to report gender, generation, or organizational level in these samples because we did not gather this data at the time these studies were conducted.

Psychological Safety and Locus of Control Study Sample (Featured in Chapter 10)

This study includes 11,294 professionals working at varying organizational levels, in the United States and elsewhere. Participants represent more than 2,000 organizations across various industries, and 88 countries overall. No one organization represents more than 5% of the sample.

Psychological Safety and Locus of Control Sample Demographics

During the period of November 2016 to July 2017, participants completed the Business Chemistry assessment online and also answered questions about how often various statements about psychological safety were true for them.

The margin error for this sample is less than +/−3 percentage points at a 95% confidence level, for all Business Chemistry types, including subtypes.

The sample for the locus of control analysis was a subset of 9,336 of these respondents.

Career Aspirations, Career Priorities, and Thriving Conditions Study Sample (Featured in Chapter 10)

This study includes 13,885 professionals working at varying organizational levels, in the United States and elsewhere. Participants represent more than 1,200 organizations across various industries, and 115 countries overall. No one organization represents more than 5% of the sample.

Sample Demographics

During the period of February 2016 to November 2016, participants completed the Business Chemistry assessment online and also answered questions about their career aspirations, career priorities, and the working conditions under which they thrive. For each topic, respondents were asked to select their top 3 options out of a list of 10.

The margin error for this sample is less than +/−2 percentage points at a 95% confidence level, for all Business Chemistry types, including subtypes.

C-Suite Study Sample (Featured in Chapter 11)

Our sample of CxOs is made up of 853 U.S.-based executives who have self-identified as holding a C-suite role in an organization of more than 100 employees. Most of them serve moderate to large organizations. More than 500 organizations are represented overall, with a relatively even distribution across industries.

Each participant took the Business Chemistry assessment online between the periods of October 2012, and July 2017. Participants and their roles were validated using external information sources such as LinkedIn and organization websites.

The margin of error for the full sample of 853 CxOs is +/−4 percentage points at a 95% confidence level. In other words, while 36% of the CxOs in our sample were identified as Pioneers, a four-point margin of error means we can be reasonably confident that, were we able to assess a large swath of C-suite leaders in the United States, the percentage of Pioneers would be somewhere between 32% and 40%. Likewise, the range for the percentage of Drivers would be 25% to 33%, for Guardians would be 14% to 22%, and for Integrators would be 13% to 21%.

Margins of error are higher for the various sub-samples of CxOs categorized by function, organization size, and gender.

C-suite Study Sample Demographics

Because the sample included a disproportionate number of men compared to women, and of CIOs compared to other C-suite roles, we explored how the representation of the Business Chemistry types in the C-suite might be impacted by a more equal distribution between women and men and across CxO roles. Weighting scores to reflect such distributions resulted in slight changes in the proportions of Business Chemistry types, but the overall representation was similar, with Pioneers being most common, followed by Drivers.

Adjusted for Overrepresentation of Position and/or Gender

Percent of Group Represented by Each Business Chemistry Type

Quotes

More than 900 Deloitte professionals are trained and certified to deliver Business Chemistry sessions within Deloitte or with our clients. Officially known as our certified network, they are more like Business Chemistry black belts, having earned their stripes with plenty of real session facilitation. Many of the quotes we offer in this book are drawn from a study we conducted with 140 certified facilitators who completed the Business Chemistry assessment online and, in September of 2015, answered a series of open-ended questions about perceptions of the Business Chemistry types. Other quotes come from interviews we've conducted and conversations we've had with executives as described below.

Facilitated Interactions

As part of our broader Deloitte Greenhouse activities, our team works with thousands of executives each year in thoughtfully crafted immersive experiences we call Labs. These Labs cover a range of topics geared to address our clients' toughest challenges; one component of which is often the dynamic between individuals, within a team, or with broader stakeholder groups. These Labs give us the opportunity to apply, refine, and extend our core research studies both through observation and through direct interviews and feedback. In addition, many of our clients have themselves adopted Business Chemistry for use across their organizations, and work with us to continue to gather insights.

Dimension Correlations

Let's start with the question of correlations between dimensions because this one is particularly easy. Eigen analysis produces orthogonal dimensions, which means dimensions that are not correlated with one another. In other words, while having a high Pioneer score means, by default, that your Guardian score will be low (because they're opposite ends of one dimension) your Pioneer score won't affect your Driver or Integrator scores since the two dimensions are independent. So you might have a high Pioneer score along with a high Driver score and a low Integrator score. Or you might have a high Pioneer score along with a low Driver score and a high Integrator score. Or you might have a high Pioneer score along with modest Driver and Integrator scores. Or, of course, you might not have a high Pioneer score at all.

However, things look a bit different when we consider the Integrator and Driver subtypes, because, after all, a subtype is generally related to the type it's a “sub” of. Overall Integrator scores are correlated with both Dreamer and Teamer scores, and overall Driver scores are correlated with both Scientist and Commander scores, all at approximately the same level: Pearson's r=.70. Correlations run from 0 to 1, so this is generally considered to be a relatively strong correlation.

Internal Consistency

People also sometimes ask about the correlations among individual items on the Business Chemistry assessment, a form of reliability known as internal consistency. Are all the items doing a comparable job of measuring the underlying construct they're designed to measure (in this case working style)? Are they all pulling their weight equally? We use the split-half method of measuring internal consistency, which compares scores on one half the items to scores on the other half. The Spearman-Brown coefficient is an indicator of this type of reliability, and for the Business Chemistry assessment it's .83, which is generally considered to signify a relatively high level of internal consistency.

Test-Retest Reliability

Another question we get in regard to the Business Chemistry assessment is about its overall stability, another type of reliability. In other words, if you take the Business Chemistry assessment twice, will you get the same scores and be identified as the same primary type both times?

We tested this with a sample of 1,420 individuals who completed the assessment twice, a year apart. We decided on this intervening time period because we wanted it to be short enough that people wouldn't reasonably be expected to have significantly changed their working style preferences, behaviors, and traits. We also wanted it to be long enough that people wouldn't have a clear memory of the assessment items or how they had responded to them previously. Since Business Chemistry is a self-report assessment, the reliability of scores depends in part on whether people are answering the questions honestly and through a similar lens each time. For example, if the first time you complete the assessment you're being very honest about how you think, feel, and behave, but the second time your answers are a bit more, shall we say aspirational, you may end up with different scores or be identified as a different type the second time around. While we'd like to assume everyone responds to the assessment with complete honesty, we wanted to reduce the chances that anyone might attempt to “game the system” to end up with a different result. We thought a year in between assessments was sufficient for doing so.

We used two methods for determining test-retest reliability. The first was comparing dimension percentiles on the first and second assessments using a simple correlation. Pearson's r for the two primary dimensions was .71 for the Pioneer-Guardian dimension and .72 for the Driver-Integrator dimension. In other words, people's results didn't come out exactly the same both times, but they were similar.

Then we took a look at how often respondents were identified as the same primary type when they took the assessment a second time. The answer to that was it depends. What it depends on (apart from honest responses) is how extreme a person's scores were on the first assessment. Let's unpack that a bit with an example.

Two people might have the following scores on their first assessment:

In both cases the Guardian percentile is the highest of the percentiles, so the person is identified as a Guardian. But person A, at the 85th percentile, is a more extreme Guardian than person B, at the 56th percentile. Now you might imagine that when a person responds to the assessment a second time a year later, even if they answer honestly both times, their scores still might change a little bit (we're not robots after all). On the one hand, if person A's scores change a little, they might end up in the 80th or even 75th percentile for Guardian (and maybe their other scores will change at similar levels), but they're still going to be identified as a Guardian, because it will still be their highest score. On the other hand, if person B's scores change a little they might end up in the 51st or even 46th percentile for Guardian. And if their other scores change at similar levels, this person may now be identified as another primary type.

In fact this example is a good illustration of what we actually find. (Imagine that!) The more extreme someone's scores are on the first assessment, the more likely they are to be identified as the same primary type the second time they take the assessment. This is just what we'd expect.

Putting some real numbers to this, overall 63% of people in our study were identified as the same primary type the second time they took the assessment a year later, but among those with stronger scores on their primary type (above the 80th percentile) that number was 72%. For those with even more extreme scores (above the 90th percentile), 79% were identified as the same primary type on the second assessment.

So what about the others? If they weren't identified as the same type, what type were they likely to be identified as on the second assessment? Usually, they wind up being the type that was their secondary type the first time they took the assessment; this was true in 24% of the cases overall. In our example above, if person B was identified as a different primary type the second time they took the assessment, chances are they'd be identified as a Driver (which was their secondary the first time).

Overall, in 87% of the cases, people were identified as their initial primary or secondary type the second time they took the assessment. That number rises along with more extreme scores, with 93% of those above the 90th percentile being identified as their initial primary or secondary type when they took the assessment again.

Other Signs of Integrity

At this point we'll loop back to highlight two points we made earlier that speak to the integrity of the Business Chemistry system. First, the fact that we replicated 96% of the item coefficients when we repeated the Eigen analysis with a second sample is a strong sign that the types themselves are stable. Second, the overlap with the types identified by other systems, based on different assessment items and derived in different ways, is a good sign of the same thing.

As we shared in Chapter 10, our research consistently reveals statistically significant differences between the types in most every area we explore, from experiences of stress and feelings of psychological safety to career aspirations and the conditions under which people thrive. Finding significant differences between the types in all of these areas suggests that the types themselves are meaningful categories.

Finally, we've seen that Business Chemistry has what's called face validity, which basically means that it seems to measure what it's designed to. We know this because time and again when we work with teams who have completed the assessment, we hear people say how well the system has managed to boil down to its essence who they are (at least at work). We can also see the light dawning as people find out the types of their colleagues, which often gives them a clue as to why it's easy for them to work with some while they experience conflict with others.