The Data Doktor
Volume 1, Number 1
June, 2004
Is all your salary data above average?
In Garrison Keillor’s mythical Lake Wobegon, “all the children are above average.” In compensation market data, Lake Wobegon could be closer than you think.
The Data Doktor is in—with news you can use
In this issue:
Ask the Data Doktor: Just because you paid for that salary survey doesn't mean it's always correct. In this, the inaugural issue of the Data Doktor, I respond to an inquiry about some fishy looking salary survey data in which the “average” salary was above the 75th percentile.
For you comp analysts, I also tell you where to find Excel techniques that help you be more productive so you can spend time on higher value work.
Who is the Data Doktor and what is NPKtools?
 The Data Doktor is a free bimonthly newsletter packed with useful tips, tricks, and solutions for everyday compensation analysis and planning problems. I’ll unveil the secrets of salary surveys, provide analytical insights for market pricing, and describe best practices and software tools that can help compensation management add more value to the organization.
Sound good? Sign up to receive future issues of the Data Doktor--Free.
What’s bugging you?
for issues to tackle in future issues. All questions about the management, analysis, and presentation of compensation data are welcome.
Who else might benefit from receiving this information?[forward-Let them know.]
Ask the Data Doktor
The Lake Wobegon Salary Survey
“Dear Data Doktor, We use the market average as our comparison point for market pricing. I have some market data where the market average is above the P75. Is this good data or is something fishy?”
On the NPK tools web site, you can read my complete answer to this question in our free white paper, “The Lake Wobegon Salary Survey.” But here’s the short version:
Our client was right to be curious about the market distribution below. This data was provided by a national compensation consulting firm and represents the second rung in a 4 position job ladder for the all-sample cut for 15 firms in the Minnesota market.
P25: 34.24
P50: 46.72
P75: 59.2
Market average: 65.82
Supposedly, the average for the distribution is above the P75, meaning that the average is above more than 75% of the rest of the sample. Is this mythical data?
To see how this might happen, I constructed a hypothetical market distribution that would conform to the summary values reported by the survey vendor. The shape of my test distribution suggests an explanation. (Readers: This makes much more sense if you can see it! View the full article online to see the charts.)
In my test distribution…
- The majority of the observations range between the P25 / P75
- Most of the observations / participants are below the $65,000 average
- Three high-paying firms are enough to skew the reported market average
- Pay for those three firms range between $120,000 and $190,000
Converting the cumulative pay distribution to a histogram, it appears that the distribution for this sample might be bimodal, that is, it has two “humps.” This might lead us to suspect that the sample may be mixing different jobs or markets that are not comparable. Bimodal distributions can also be an indicator of incorrect job matching. Further investigation suggested our client’s market data was misleading due to a weaknesses in vendor’s methodology for survey job matching.
Isn’t this an unusual quirk?
Not really. In our practice over the last two years, we’ve found that reported market averages exceeded P75 or were less than P25 for 2.5% of the 180,000 records we analyzed or about one out of every 40 market records. However, in one survey that provides a wide variety of market cuts, the incidence was 11%. In this case, more than one out of ten times the data suggested a potential bimodal risk.
So how do you know if your data is wobegone? And what do you do about it?
The answers are a little too long to go into here, but please check the full article online to get the seven specific steps for identifying and managing this data anomaly.
Analysts' tips and tricks
Managing company data. Job matching issues. Survey management. If you’re a Comp Analyst, you know that a lot of the numbers on which salary decisions are based come down to whole lot of cutting, pasting, sorting, and other numeric finaglings.
In Analysts' Tips and Tricks, we provide a detailed guide to the real nitty gritty of comp analysis, starting with the one tool almost everyone uses, Microsoft Excel.
On the NPKtools web site you’ll find information on using Excel Lookups for survey submission job mapping. If mapping data from two or more files to create a coherent and accurate data submission is a concern of yours, you won’t want to miss this practical advice.
You’ll also find a short piece on Using Excel text functions to create VLookup values.
About this Data Doktor guy
Hi, I am Lindsay Scott, managing partner of NPKtools, a compensation analysis software, consulting, and outsourcing firm. I am a former Hay consultant, where I focused on software development, information analysis, and information delivery. I have an MBA from Duke University, and before joining Hay, I consulted on energy, economic and regulatory issues as a senior management consultant for EDS, Energy Management Associates, and DRI/McGraw Hill.
I like to believe that my background as a market analyst in broader industry settings gives me a unique perspective on compensation markets. I definitely believe that understanding pay markets can be an important ingredient in company competitiveness. I hope that the Data Doktor will help compensation professionals bring better information to bear on company decision-making—and thereby raise their profile as organizational contributors.
What you can do now:
(or call me at 770-740-8660).
Recommend this page to a friend.
Check out the free resources on the NPKtools web site.
Copyright 2004, NPKtools, inc. <http://www.npktools.com> Distribution of this publication is encouraged IF reproduced in its entirety.
|
