This document provides examples of applying rules of thumb to interpret exposure monitoring data and assign exposure risk categories. It shows how to determine the median, calculate multiples of the median, and compare them to the occupational exposure limit to estimate the 95th percentile exposure and assign a category. Categories range from 1 to 4, with higher numbers indicating greater risk. The examples demonstrate assigning categories based on whether any exposures exceed limits or guidelines. The rules of thumb are meant to aid quick decision making when only limited monitoring data is available.

1. AIHA VIDEO SERIES:
MAKING ACCURATE EXPOSURE RISK
DECISIONS
Video 1B
Rules of Thumb for Interpreting
Exposure Monitoring Data
1

2. Disclaimer &
Copyright
Although the information contained in this session has been compiled from sources believed to be reliable, the presenter and AIHA®
make no guarantee as to, and assumes no responsibility for, the correctness, sufficiency, or completeness of such information.
Since standards and codes vary from one place to another, consult with your local Occupational or Environmental Health and Safety
professional to determine the current state of the art before applying what you learn from this webinar.
AIHA must ensure balance, independence, objectivity, and scientific rigor in its educational events. Instructors are expected to
disclose any significant financial interests or other relationships. The intent of this disclosure is not to prevent an instructor from
presenting, but to provide participants with information to base their own judgments. It remains up to the participant to determine
whether an instructor’s interests or relationships may influence the presentation.
Session presentation material belongs to the presenter with usage rights given to AIHA. This session and associated materials can
be reproduced, rebroadcast, or made into derivative works without express written permission.

3. Disclaimer &
Copyright
Handout Information
All AIHA University session handouts are produced by AIHA as submitted by the instructors, and in the instructor determined order.
AIHA does not change or modify handout content; we may adjust images to improve layout and clarity.

4. AGENDA
•Review: Key points from Video 1A
•Rules of Thumb
•Examples
•Key Resources

5. CLICK TO EDIT MASTER TITLE
STYLE
5
QUICK REVIEW . . .

6. Effective and Efficient
Exposure Risk Management
Effective:
Ensure that no worker has
unacceptable exposures
Efficient:
Do it for minimum cost
6

7. 7
7
What if our exposure assessment is wrong?
If we underestimate the exposure?
• Increased risk to employees
If we overestimate the exposure?
• Unnecessary constraints for employees
and production
• Unnecessary expenditures for controls
7
Well-Designed Exposure Risk
Management Strategy
We Want: We Don’t Want:
Good Data Bad Data
To Be Effective To Not Be Protective
To Be Efficient To Waste Resources
No Biases Biases (High or Low)
Low Uncertainty High Uncertainty
Correct Decisions Wrong Decisions

8. Decision Statistic:
“Strive for at least 95% confidence that the
true 95th percentile is less than the OEL”
OEL
95%ile
95%ile UCL
0 1 2 3 4
-2
0
2
4
6
8
10
0 0.2 0.4 0.6 0.8 1 1.2
Best Estimate
95% Upper Confidence
Best Estimate
Exposure Profile
95% Upper Confidence
Exposure Profile
8

9. AIHA Exposure Rating and Control Categories
Increase Effectiveness and Efficiency
• Avoid diminishing returns from
“over-refining” exposure estimates
• Streamline Documentation
• Facilitate Qualitative Exposure
Judgements
• Drive consistent follow-up
management and control activities
which lead to consistent risk
management.
9

10. ?
Exposure Risk Decisions:
How Accurate Are We?
** Decision statistic = 95th percentile
Sample Results
(ppm)
18
15
5
8
12
When We Have Monitoring Data . . .
10

11. Judgement Accuracy is Poor If We Don’t Use
Statistical Tools When We Have Monitoring Data

12. Lack of Familiarity with Properties of the
Upper Tail of Lognormal Distributions
95%ile
0
1
2
3
4
5
6
7
8
9
0 0.1 0.2 0.3 0.4 0.5 0.6

13. Lack of Familiarity with Properties of the
Upper Tail of Lognormal Distributions
95%ile
0
1
2
3
4
5
6
7
8
9
0 0.1 0.2 0.3 0.4 0.5 0.6
• Skewed to high end
• Unlikely to have result in
upper 5%ile when
number of samples is low

14. CLICK TO EDIT MASTER TITLE
STYLE
14
RULES OF
THUMB

15. Rules-of-Thumb to Aid Data Interpretation
Given:
• GM = median
• X0.95=GM x GSD1.645
… Rules-of-thumb, or
guidelines, can be devised
for quickly estimating
from limited data the
range in which the true
95th percentile might lie.

16. Rules-of-Thumb to Aid Data Interpretation
Given:
• GM = median
• X0.95=GM x GSD1.645
… Rules-of-thumb, or
guidelines, can be devised
for quickly estimating
from limited data the
range in which the true
95th percentile might lie.
GSD
Multiple of GM (median) to
Calculate 95%ile
1.5 1.95
2.0 3.13
2.5 4.51
3.0 6.09
16

17. Rules-of-Thumb to Aid Data Interpretation
Given:
• GM = median
• X0.95=GM x GSD1.645
… Rules-of-thumb, or
guidelines, can be devised
for quickly estimating
from limited data the
range in which the true
95th percentile might lie.
GSD
Multiple of GM (median) to
Calculate 95%ile
1.5 1.95
2.0 3.13
2.5 4.51
3.0 6.09
17
4
6
2
Low
High
Variability

18. Rules of Thumb to Aid Data Interpretation
• For Low n: If any measurement > OEL then Category 4: 95%ile > OEL
• Determine Median of the Data
• Calculate and compare to OEL:
2 x Median
4 x Median
6 x Median
• Emphasis on 2 x Median if
data have little spread
• Emphasis on 6 x Median if
data have large spread
Note: A lower category is not an option if
any measurements are in a higher category.
Variability ROT
Multiplier
Low 2
Medium 4
High 6
18

19. CLICK TO EDIT MASTER TITLE
STYLE
19
EXAMPLES

20. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 30, 17, 7, 13 , 63, 5
Approximate X0.95
20

21. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63
Approximate X0.95
21

22. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15
Approximate X0.95
22

23. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90
Approximate X0.95
23

24. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90
Approximate X0.95
24
One measurement > 50%OEL (in Category 3). This
eliminates Categories 0, 1, and 2.

25. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
Approximate X0.95
25
One measurement > 50%OEL (in Category 3). This
eliminates Categories 0, 1, and 2.

26. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
Approximate X0.95
26
OEL = 100 ppm GM = 15.6 ppm
GSD = 2.54
95%ile = 72.6 ppm
95%ile UCL95,95 = 497 ppm
Sample Results
(ppm)
30
17
7
13
63
5

27. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6
Approximate X0.95
27

28. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6
Approximate X0.95
28

29. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36
Approximate X0.95
29

30. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
Approximate X0.95
30

31. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
Approximate X0.95
31
Sample Results
(ppm)
6
OEL = 100 ppm GM = 6 ppm
GSD = ????
95%ile = ????
95%ile UCL95,95 = ????

32. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
Approximate X0.95
32
Sample Results
(ppm)
6
OEL = 100 ppm GM = 6 ppm
GSD = ????
95%ile = ????
95%ile UCL95,95 = ????

33. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 33, 37, 9, 109, 8, 5
Approximate X0.95
33

34. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109
Approximate X0.95
34

35. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109
Approximate X0.95
35
One measurement > OEL so the ROT rating is Category 4.

36. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109 4
Approximate X0.95
36
One measurement > OEL so the ROT rating is Category 4.

37. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109 21 42 84 126 4
Approximate X0.95
37
One measurement > OEL so the ROT rating is Category 4.

38. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109 21 42 84 126 4
Approximate X0.95
38
Sample Results
(ppm)
33
37
9
109
8
5
OEL = 100 ppm GM = 19.1 ppm
GSD = 3.23
95%ile = 131 ppm
95%ile UCL95,95 = 1479 ppm

39. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109 21 42 84 126 4
Approximate X0.95
39
Sample Results
(ppm)
33
37
9
109
8
5
OEL = 100 ppm GM = 19.1 ppm
GSD = 3.23
95%ile = 131 ppm
95%ile UCL95,95 = 1479 ppm

40. Rules-of-Thumb Examples
Data
Set
Data (ppm)
(OEL=100) Median 2x 4x 6x
Exposure
Category
(1-4)
A 5, 7, 13, 17, 30, 63* 15 30 60 90 3
B 6 6 12 24 36 2
C 5, 8, 9, 33, 37, 109** 21 42 84 126 4
D 3, 5, 12†, 20† 8.5 17 34 51 2
E 78 78 156 312 468 4
F 1, 3 2 4 8 12 1
G 17†, 18†, 31†, 45† 24.5 49 98 147 3
H 4, 5, 6, 12†, 14†, 36† 9 18 36 54 2
Approximate X0.95
40
* Dataset A: one measurement > 50%OEL. This eliminates Categories 0, 1, and 2.
**Dataset C: one measurement > OEL so the ROT rating is Category 4. (The sample 95th percentile is 131.)
† Datasets D, G, and H has measurements > 10%OEL. This eliminates Categories 0 and 1.

41. Rules of Thumb Comments
• The Rules of Thumb help us understand the implications of the upper
percentile decision statistic as opposed to some measure of central tendency.
• The Rules of Thumb counter our human world view which tends to think
symmetrically and underestimate the extent to which the 95%ile of a
lognormal distribution is likely to exceed the median sample result, even at
relatively low variability.
• The Rules of Thumb can be applied to censored datasets – i.e., datasets
containing non-detects – when the NDs are all less than the median.
• While the Rules of Thumb do improve judgments, the ultimate answer to
improved accuracy is to use statistical tools when we have monitoring data.
41

42. Use Statistical Tools!!
42
95%ile = 1.2
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 0.5 1.0 1.5 2.0
Concentration (mg/M3)
UTL95%,95% =
16 mg/M3
AIHA
IHSTAT
™
Industrial Hygiene Statistics Beta 0.9 - For trial and testing only - Please do not distribute
Data Description: John Mulhausen
OEL DESCRIPTIVE STATISTICS
5 Number of Samples (n) 15
Maximum (max) 5.5
Sample Data Minimum (min) 1.2
(max n=50) Range 4.3
No less-than (<) Percent above OEL (%>OEL) 6.667
or greater-than (>) Mean 2.680
1.3 Median 2.500
1.8 Standard Deviation (s) 1.138
1.2 Mean of Log (LN) Transformed Data 0.908
4.5 Std Deviation of Log (LN) Transformed Data 0.407
2 Geometric Mean (GM) 2.479
2.1 Geometric Standard Deviation (GSD) 1.502
5.5
2.2 TEST FOR DISTRIBUTION FIT
3 W Test of Log (LN) Transformed Data 0.974
2.4 Lognormal (a=0.05)? Yes
2.5
2.5 W Test of Data 0.904
3.5 Normal (a=0.05)? Yes
2.8
2.9 LOGNORMAL PARAMETRIC STATISTICS
Estimated Arithmetic Mean - MVUE 2.677
1,95%LCL - Land's "Exact" 2.257
1,95%UCL - Land's "Exact" 3.327
95th Percentile 4.843
Upper Tolerance Limit (95%, 95%) 7.046
Percent Exceeding OEL (% > OEL) 4.241
1,95% LCL % > OEL 0.855
1,95% UCL % > OEL 15.271
NORMAL PARAMETRIC STATISTICS
Mean 2.680
1,95%LCL - t stats 2.162
1,95%UCL- t stats 3.198
95th Percentile - Z 4.553
Upper Tolerance Limit (95%, 95%) 5.60
Percent Exceeding OEL (% > OEL) 2.078
Linear Probability Plot and Least Squares
Best Fit Line
1%
2%
5%
10%
16%
25%
50%
75%
84%
90%
95%
98%
99%
-5 0 5 10
Concentration
Log-Probability Plot and Least Squares Best Fit Line
1%
2%
5%
10%
16%
25%
50%
75%
84%
90%
95%
98%
99%
0 1 10
Concentration
Idealized Lognormal Distribution
AM and CI's 95%ile
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0 1 2 3 4 5 6 7
Concentration
Sequential Data Plot
0
1
2
3
4
5
6
0 2 4 6 8 10 12 14 16
Sample Number
Concentration
IH
Data
Analyst
Exposure Rating Category
<1%OEL <10% OEL 10 – 50% 50 – 100% >100% OEL
Probability
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0
0.087
0.4
0.513
OEL
Likelihood that
95%ile falls into
indicated
Exposure Rating
Category
Initial
Qualitative
Assessment
or Validated
Model
Prior
Exposure Rating
0 1 2 3 4
Decision
Probability
1
0.8
0.6
0.4
0.2
0
0.05
0.2
0.5
0.2
0.05
Monitoring
Results
Likelihood
Exposure Rating
0 1 2 3 4
Decision
Probability
1
0.8
0.6
0.4
0.2
0
0 0 0.06
0.376
0.564
Integrated
Exposure
Assessment
Posterior
Exposure Rating
0 1 2 3 4
Decision
Probability
1
0.8
0.6
0.4
0.2
0
0 0
0.225
0.564
0.211
Expostats
™
Traditional Statistics Bayesian Statistics
42

43. CLICK TO EDIT MASTER TITLE
STYLE
43
KEY RESOURCES

44. Free IH Statistical Analysis Tools
• AIHA IHSTAT - Excel application that calculates various exposure
statistics, performs goodness of fit tests, and graphs exposure data.
https://www.aiha.org/public-resources/consumer-resources/topics-of-interest/ih-apps-tools
• Bayesian Decision Analysis – IH Data Analyst Student Edition
Computer Tool
https://www.easinc.co/
• Expostats Bayesian IH Data Analyst Tool
http://expostats.ca/site/en/index.html
44
FREE
TOOLS!!

45. Free Exposure Assessment Tools
• IH/OEHS Exposure Scenario Tool (IHEST)™
Excel tool to aid Basic Characterization
• IHSkinPerm™
Excel tool for estimating dermal absorption.
• Basic Exposure Assessment and Sampling Spreadsheet™
Excel template for entering EA/BC and sampling data
• The Qualitative Exposure Assessment Checklist (The Checklist)™
• IHMOD 2.0™
Excel-based mathematical modeling spreadsheet
https://www.aiha.org/public-resources/consumer-resources/topics-of-interest/ih-apps-tools
45
MORE
FREE
TOOLS!!

46. Exposure Decision Analysis:
Competency Assessment
Exposure Decision Criteria
• Allowable Exceedance
• Needed Confidence
• Use of Exposure Categories
Traditional Industrial Hygiene Stats
• Properties of a lognormal distribution
• Upper percentile estimate calculation & interpretation
• Tolerance Limit calculation & interpretation
Bayesian Decision Analysis (BDA)
• Properties of a lognormal distribution
• Upper percentile estimate calculation & interpretation
• Tolerance Limit calculation & interpretation
Data and Similar Exposure Groups (SEGs)
• Rules for combining data
• Indications that a SEG may need refining
Decision Heuristics and Human Biases
• ​Common sources of bias in data interpretation and
exposure assessment
• How to avoid bias in data interpretation
Exposure Data Interpretation
• ​Most likely exposure category given data
• Meet the certainty requirement given data
Techniques for Improving Professional Judgments
• ​Feedback loops (quantitative judgment > monitoring >
qualitative judgment)
• Group judgment sessions
• Documentation of rationale
• Break decisions into aggregate parts (Modeling)
46

47. Learn More:
aiha.org | 47
• Papers:
• Logan P., G. Ramachandran, J. Mulhausen, S. Banerjee, and P. Hewett “Desktop Study of
Occupational Exposure Judgments: Do Education and Experience Influence Accuracy?” Journal of
Occupational and Environmental Hygiene, 8:12, 746-758, 2011.
• Logan P., G. Ramachandran, J. Mulhausen, and P. Hewett:” Occupational Exposure Decisions: Can
Limited Data Interpretation Training Help Improve Accuracy?” Annals of Occupational Hygiene, Vol.
53, No. 4, pp. 311–324, 2009.
• Vadali, M. G. Ramachandran, J. Mulhausen, S. Banerjee, "Effect of Training on Exposure Judgment
Accuracy of Industrial Hygienists”. Journal of Occupational & Environmental Hygiene. 9: 242–256,
2012.
• Arnold S., M. Stenzel, D. Drolet, G. Ramachandran; Journal of Occupational and Environmental
Hygiene, 13, 159-168, 2016
• Hewett, P., Logan, P., Mulhausen, J., Ramachandran, G., and Banerjee, S.: “Rating Exposure Control
using Bayesian Decision Analysis”, Journal of Occupational and Environmental Hygiene, 3: 568–
581, 2006
• Jérôme Lavoué, Lawrence Joseph, Peter Knott, Hugh Davies, France Labrèche, Frédéric Clerc,
Gautier Mater, Tracy Kirkham, “Expostats: A Bayesian Toolkit to Aid the Interpretation of
Occupational Exposure Measurements”, Annals of Work Exposures and Health, Volume 63, Issue 3,
April 2019, Pages 267–279

48. 48
• Books:
• A Strategy for Assessing and Managing Occupational Exposures. 4th Ed.
AIHA Press. 2015.
• Opinion:
• Mulhausen, J. “Faulty Judgment” President’s Message. The Synergist.
(November 2021).
• Mulhausen, J. “How to Improve Exposure Judgments” President’s
Message. The Synergist. (December 2021).
• Videos:
• Mulhausen, J. “Top 10 Imperatives for the AIHA Exposure Risk
Management Process.”
Free from AIHA at https://online-
ams.aiha.org/amsssa/ecssashop.show_product_detail?p_mode=detail&p
_product_serno=2650&p_cust_id=&p_order_serno=&p_promo_cd=&p_p
rice_cd=&p_category_id=&p_session_serno=72069269&p_trans_ty=
Learn More:

49. AIHA VIDEO SERIES:
MAKING ACCURATE EXPOSURE RISK DECISIONS
Video 1A: Exposure Variability and the Importance of Using Statistics to
Improve Judgements
Video 1B: Rules of Thumb for Interpreting Exposure Monitoring Data
Video 2: Introduction to Bayesian Statistical Approaches and Their
Advantages
Video 3A: Free Bayesian Statistical Tools: IHDA Student Edition
Video 3B: Free Bayesian Statistical Tools: Expostats
Video 4: Implementing AIHA Strategy Using Statistical Tools: Examples
49
Join us for the next video in the series . . .

50. AIHA VIDEO SERIES:
MAKING ACCURATE EXPOSURE RISK
DECISIONS
Video 1B
Rules of Thumb for Interpreting
Exposure Monitoring Data
50