.h...+,,,_ _. . .... . ...flwqu..§a1
w. ..

   

              

 

..t._m-.w:£“wo.,..... ..L..\.c.- fi .14." ‘ ...I. o
Kr%;..%¢m§ an . r: 3.. ...-wanna...” .. ...-m . ... . .

. . . 1 a...» ......i : .-....um-u. 54-. .2: -..-anew?

.. ..fi 13; .. 4......- ilflw... :6: . . ....u- rum-M.

.mmw...-.-......ww§a.=.w, ...-m. y...- n. i. - .

. quwfiafifinfir we. _ .-.-mug.

. gown-Wu ...-km».- , fist-13.”...

...TT

      

   

     

              

        

    

         

    

             

              

. . . ... z . ...:phk
... ...... 3%? _ .

...?ik... ...-.... gnaw...
a... .... .... w...
' "X'

. Fa
r- "a...
2%....fi
. . m. .11. Ana-v...
. ._ N5... ....-

 

v-

I.

v!
_

     

u

 

I"
...
-

. ...-#5....- .31.-

I- .

a . uni... uwyfimwan ..-
. . ...1 - .1.
Mg,- 111..-;
.....- . .

. .9"!

   

 

     

. .. ...fialfliifi ...
» .IKIILH-Rrv 3!.de
4.5.5- NFL-nan...”-
. - DC.‘ Oil-rt Iv. .9.
... . - .-

. - luau-«1:5? Afl$ .
arr-n ,.. .fi-k-WW-E-pfl: .FS
: . ..vflw. ..n-N ...mW-h-hw-xfi...
. Ada-4mm? ....-.u......,. nwame-fififidflfi.

u.

   

  

ll ...-.11.
.3. Lip)

                   

               

        

 
     

          
  

 

.. . n. . <
. . flfin ... ....-...-.-. - . S had....9-uui-§.fi-1Pv.1.c .- . .. ._
. .u- ...-..-3.r.u..-..r-..-- 95.41%. a... .65.“an .... Cf...»- .. . . . .
. z --.. v ... . hi... ......ub....4t-3..3 ......z. .. ....EJ-u-F-hv- t .l- tit-hu- !
..a-u-mfim 5.9....»- nmhvdv ......M bun-...- ..-....- I. Sum-u... L 52.3- 69.3“. . . hum. ..Mu..rivn--Rz ,-
ai-C:-f;a..1-.-4ul-.. u..- --.-..-- . $.24.- . QM"... 2- Gun-- 5...... in... 511)., fi. . 33- Ir! -
4dr? o..Lu»-.Ipv.o.....3 .\ ...-lawmflnndfixc 3.321... W. hahffian ”Gui...- -u..l..¢..h.l-Ys.?.r§..rc .

       

‘s-0 .1 I I- O. ‘A I O ...-I. 5 l'hIQI-u.‘l .‘I
.Ma-wuflmhuwm-Wm. - ..wmodmmflxi-a-un - £48... .. -.:..r.. - . -

  

.. then-.... .... c.13- to.

    

 

 

       
                  

           

     
                     

  
     

     

 

                  

    

               

   

   
    

         
              

       

       

  

   

      
          

 

 

   

      

-
. . =25; . ..3. . .9... fl. . ...-115.3... (-91-.
- ... . . .v. . . .12? 1d? . 15-5 ... .5 2-19., -. .
i..-h-u...-.In.£m-.u- ...... ...-h -.. .Nn..fi-H-..;r a gun-nu-w-«xs-iqiéi; .. - . .nb .... . z
- -. 1...-....n..-...(x - ...-J. a ...-..- ..1 k... .. «1L..- wz-a- . . , .. -.. ...-.... .1...
)7 $lV-J. l . 1L4 In"! (nib! VI~vI 501.11 1 ...-...f- .- .I?..l01h(.u , 4 ....Qk'.tv“dvc\10 ‘14». pgfiflka .0!!!-
. . 1.- . - . - . .. .. 1., .11.. ..i- . ,l ..... #3 ..I
3 «....o. ,. .34- .m- -.. -.. ...» .7. . ...-3 . .. ...-ht. .1- ui- -.. :- .
. . ...-.....skdnl__i.-. 1. - 5...... ...d . 1.5.4... 189-... .- ...-...“. -..-i: . Jun: -1; :10 . 1...... .1 4!- . ...-3...
. . .... Ianl!‘ . P \. ... I- tn :fii-I-Ifw rt! 3!...- ..I u!.. .. ..., ’1'.» 1c- Az-bvaa :- . . trot ..l.
.14 .-I . I... ..- Eli-1 «é. y - ...-.31... .- 93-..- -. _ - ..-.(..n-.- .271... , .. . {ah-7!.- «.-
..... t {...- ...-...... ... 1.4%....- .-.-.-...I.-- . .. 2... 413.311.. . 4 art-It. I! . ... .
.. v‘ll v3 v1 l1. .. l. .. 'i I ...-C II I '- v..t-t.lIl-qv . .11. ~ I. r XII-il‘lt-r... -I-A2I In...» 111...: IS- 911'
.11.. . I ’1 ill?! c IIIII I a #t..vv%t..l.~! - -- 1.1.13.1.- . .-I1Ivn~4..lul| p all '1 t» id'-
. -- - . ,- .- ..-; ...-2 1:. $5 1- :...i. . . 1.1-... . {Ii-11¢ :- - . 9-
..rklr; flvr6.‘§? ., AVE-l. .. I. IDInWOer‘il ain‘t utilvl‘vtfifiu‘l» tau“ . . .. wovtxz-Ol-DuOu \ :r’lS‘lu . v . O
.. r-Ll-huu. l.- hfihfir 6..........F.c:....v..--- 4...“... ...-...r. . . §.1-.-.1-Ivl§1-l- - (nonna- ii- 1.-
.2'.l-Iv ..- ..... 11 n'....... 11-36..- . :II :1: - .lI ......... ll! 11-1- P. In! 1 L."
Italvl-ll-‘Lr. w‘l‘v .. 1!... 11:7 . V.- v . . vol-l A:
-.¢......mu-..v#-.--..--:u.-nv ...-$1...- .Jfiuu... . . . {lbw-m1}... .... 1....-
-. .-. v F:§mflh§h-¥§ ,- 1.1-;- . . . .. . . oil-io- tl'lv). . tfiufldur-i
inn-u, 5-1.1}: 5-1.: ... a. . {flu-.- .. .. . . . 313-.“- é-lvr-v [...-.1
... i- at It". a A ...-v. 2...? i-rI .1 '. . V . . . . -.. . - . . . . 7|,1‘0-I'A a. D‘l.
. tic-III: I Elli-... ..lc- .. . . .. 5.. - .tl‘u-v .- - . I . 51!"; v.
. .... ... .... .31.. ...- ...f- , - - ...-1.3- . 1 -rv- - Mann-11v.
...- . - . .1 -v 3- - -..-...: .. . . .. -
ali‘r-VW - L0.Din).t . 1.7.1 - .(-x~v - . .. y a unit-<1 ‘90.- Iii-u
«1...! - - .---...--- <1. . . . 31... . . .. 1..., .
i I III"! 3’. V. IV. 0 - - u i u . Sll-tO'U'OL ’1 ‘l l
. .g. ...-...... .y... . .. . . r . . . : ., . . . 1113-1-11.-
. . \- .. ...-n 1......- 1-.-.. .. , . . . . . _ 91:. -
. ,.-.o..lu);-I)Dbln- ...I n at . ~ . -. . . . . . . , . {-r ..l- IIVJON-v-Jf
blx . ......” 141‘ Hal-:59)?- o . J . - I . z r . . -. II. n ’u If I
... )1 311-0 t-pa..l 57‘ a £9.15 7- l. . . . . - . . . J1\VlI'-ut 1- l.- IYII. 1' . b
115...»..- u - .... ...-315.. (.... .. . . . . .11..- I..-.t.-. .r-i. Lilia-Hull .
. 5-. ..l-nu- ...-v- . a... 5! 92:12.... . . . . . . , : . . 1.2115111413131111. 1..
. . I.” Lat-fnuFU-‘zl 34.1.3111?! ark-unal- . . . . . . l I}: - . 3. 1111417.... -
. ,. . .3- .. .. i .. .. - 1.27.37 . . . .. .
....o.!-.1.-:-l--.fl..-si1. :- ol! . .
. . ~ [dunk-2.3!: 1- ...... . . .
Vii-312.1. n. 19.5.... - .- ..h-l-(l!
...H-ulslmunui... {...-1.1». . .
I.Ia‘:§ i n . v.1 I l\.KOa 9|-
1- ,623 31-1.! h...»- --
. 112.32.81.1- . {IA-.1- 31.- .- .. -..--.....--.
th‘l II ‘19.“. .... 1": ’c?l, I‘M-Ehnuo’tv It). .I I ~ V In" 3"3‘hl-}~ all
33-1 .v 1.- - .-. .. .f ,z.. .2... . 2-. . - - - c- . ..-
...... .- . . I ..- . -. . 5-1;! - . . -
. --.€-o....Jn.-IA:-.--.-.1u-..“a ...”.rma-‘yfi-ann-ffi -.-fi--.-£u-u-.-om-J..-§.r.--m5mw-
{41.11215} inn-1411;. 1-...1-8-1Qt-ui-s-lz.“...... llWI-1711-112i-1- .-- .
v I - o . 7:? D- ‘14): vugfl: )t‘II.-ul-' liO‘
5-3.3:- . 9.5 b -

  

u.- .. .. {...-.33. - ...-1.01.. .
VB; I'GV-Ilhvl I! like.“ - plut' Judah“. Y5 I|I. 5‘.- [‘1‘ lb‘

. l- ...: 1 (- - -.

. . ......- ...-...?95-31311 ; an... ...-.r
111.13. .. . .95.-

 

I

 

 

flulfiluhhffl. "Navy ..
:l-un-l - ...Ifl'hnvI-v-J-

urn.
It}

 

:o-l
. 1| - 7| A $2.11!! 1,-1.-

 

    

 

      

 

         

. .u... .‘vuxlew («J-7
w . ..- JJH mall-.... .-
V y? Ab’OV'tt‘l l! 1 II .| lnzl‘t.‘ 07-1
H1 14. . .--.D‘t|rfi||N-anhll.- “AT-40.11, ..1}
41.1 lit-1v- 11: 1-11-11: ‘-
. .-.. t ........ 11-11:}-
.. o..u-a . .
21-0-1.
«-

 

II»- ..
- Uhnv‘vh-n-WHD

 

   

- .. . A LIIi-x- ‘lll-sl"-’1ll11l‘l1
-.-| ’1 0- . l ...-1.11! .ll..l1lf fl.“ I ll.)
- 1.1-0... I.M-v-.-l-r .4 I. .)||-.1 v1.1! '1 A. I

  

 

    

r»! In ~If1-II1' “'VHIIII‘ Ill-1.14!- u‘l’
..vIIOV-VIJI- lulu-1141111111.,Ol 11'. Io!)|-Hl |OOL
vl1- .v-..-. 0101‘!- l'.‘ $111!.r I!
n; oft-d...- 199 1 l1! lit-Int" Jittl
1!- I.- - f-.. |- 1- 1 la] 1111-11 1
- lrtl l.- ..F 0 111I1.| {ll 1 -'!ll1-: lv-[II 111
-1«.\.- 1-114 -J| 011-01111) 111.32-111.11-1
I 17' nu-r......O--. 7...D-1---I|11-61 - l-1|-I1‘|l 01- 1
l- l.-v-1 V 1.‘ . I 2 .I.vl-l-II. .111). I‘ll-lf-II-r’ le’fa’
. {LI-11 1| ¥.l A! ~.v«-V.l-11-l-|III ‘i-ulli‘l
. . .... FJ-f..1- . .. {LI-.... {If-1410111111111
3r . {-1 11.1 a... 10-1.! .-.l|1-?1-I|-F|-. 111-!) -l-
. . . . A - ...-1:. of- . -- i1 Ill-(1.1; v 1|.llill’ll’ la
.....ll-1«-.-1. . 111- l..- .....- . -.-1-clt11‘...-IL-l-l -1 1 .
..t‘ -!.| o.l'l‘l’il {til-v - VI- 1 - - - V . . ‘1' ||
- up: -‘ | .u i v 7- L-

 

'HESlS

SlTYL LBI RABIES

llllllllll‘llllllllllllllllllllllllllllllllllll| ll

3 1293 01570

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled
AN ASSESSMENT OF RISK PERCEPTION IN

TRANSPORTING HAZARDOUS MATERIALS BY RAIL
THROUGH DEXTER VILLAGE, MICHIGAN. AND

”Imsmfi‘ééeé’t‘éfiooy

BETHANNMYERS

has been accepted towards fulfillment
of the requirements for

M-A- degree in Geoqraphy

 

Wei/iv 233333;,

Major NU essor

Date EKKQO [C] r)

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Mlchlgan State
‘ Unlverslty

 

 

 

‘ PLACE ll RETURN BOX to roman this chockout from your record.
. TO AVOID FINES Mum on or before date duo.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

l_
- -
[:r—J

I ISL—J
l___Jl——l-
L____

l—W—‘T—j

MSU Is An Affirmative Action/Equal Opportunlty lnstltwon

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AN ASSESSMENT OF RISK PERCEPTION
DNTFRAUQSPCHITHNCI
HAZARDOUS MATERIALS BY RAIL
TTflRCflJCfli
IHiXTIflIVHLlJUEE,LflKHIKlAbI
3AA“)
thMNHSEKHfl83CHIK)

by

Beth Ann Myers

A THESIS

Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Geography

1 997

ABSTRACT
AN ASSESSMENT OF RISK PERCEPTION
1N TRANSPORTING
HAZARDOUS MATERIALS BY RAIL
THROUGH
DEXTER VILLAGE, MICHIGAN
AND
MIAMISBURG, OHIO
by

Beth Ann Myers

Recognition of hazardous materials in manufacturing and recognition of the
disposal risks has been acknowledged in past literature, but little research has addressed
the intermediary step of hazardous materials transportation. This thesis is concerned with
identifying risk perceptions in conjunction with the transportation of hazardous materials
by rail. A survey questionnaire was distributed in the study areas, Dexter Village,
Michigan, and Miamisburg, Ohio, and then evaluated for risk perception levels and
demographic characteristics of the respondents. Statistical analysis including regression
and t-tests were incorporated to ascertain whether certain demographic variables can be
indicative of one's risk perception level. The major finding of this study, which parallels
past perception literature, is the higher risk perception levels of female respondents. Other
comparisons between the two study areas and the remaining demographic variables were

inconclusive.

ACKNOWLEDGEMENTS

I would like to thank the following people for making my Master’s experience
rewarding and this thesis a lesson in tenacity. First of all, Dr. Pigozzi for his
encouragement, patience, and unwavering support throughout this process. Dr. Burley
and Dr. Olson for their comments, suggestions, and help with the final thesis. My
husband, Jeffrey, for his help in all phases of this thesis. And lastly, my family for their

continued encouragement and support during my tenure at Michigan State University.

iii

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
INTRODUCTION AND BACKGROUND
Structure of this Paper
RELEVANT LITERATURE
The New Environmental Paradigm
Risk and Experts vs. Laypeople
Factors Influencing Risk Perception
Perceptions Concerning Transporting Hazardous Materials
Regulations
METHODS
Research Design
Study Area
Miamisburg, Ohio
Dexter Village, Michigan
Survey Instrument
Pre-Test
Survey Distribution

Analytical Methods

iv

page

viii

10
12
12
13
15
19
22
23
23
24
26
27
31
33
34

36

ANALYSIS
Risk Perception
Demographic Profile
Demographic Issues
Hypotheses Testing
DISCUSSION
Conclusions and Recommendations
General Environmental Issues
Transportation Risk Factors
Site Risk Factors
Hypotheses
Structure of the Study
The Survey
Questionnaire Structure
APPENDIX A
APPENDIX B
APPENDIX C
APPENDIX D

LIST OF REFERENCES

page
37

37

42

43

59

61

61

62

63

65

66

67

73

74

75

78

80

LIST OF TABLES

page
1. USEPA Survey of Public's Risks .......................................................... 16
2. Socioeconomic Profile of Case Study Areas ........................................ 35
3. Summary of Variables ........................................................................... 38
4. Mean Perception Scores for General Environmental Issues ................. 4O
5. Mean Perception Scores for Transportation Risk Factors .................... 41
6. Mean Perception Scores for Site Risk Factors ........................................ 42
7. Demographic Profile of Survey Respondents ....................................... 44
8. Bivariate Correlation Coefficients of LENGTH, EDUC, PROXIMITY 45
9. Bivariate Correlation Coefficients of ENVIRO, TRANS, and SITE 46
10. Difference of Means T—Test between Dexter Village and Miamisburg 47
11. Difference of Means T-Test between Female and Male Respondents 48
12. Regression of TOTAL and EDUC ........................................................ 49
13. Regression of TOTAL, EDUC, and TOWN ......................................... 50
14. Regression of TOTAL, EDUC, TOWN and ET ................................... 51
15. Regression of TOTAL and PROXIMITY ............................................ 52
16. Regression of TOTAL, PROXIMITY, and TOWN .............................. 53
17. Regression of TOTAL, PROXIMITY, TOWN, and PT ....................... 54
18. Regression of TOTAL and LENGTH .................................................. SS
19. Regression of TOTAL, LENGTH, and TOWN ................................... 56
20. Regression of TOTAL, LENGTH, TOWN, and LT ............................. 58

- page
21. Review of Hypothesis Outcomes ........................................................... 58
22. Sample Of Reconstructed Survey ......................................................... 7O

vii

LIST OF FIGURES

. Case Study Areas Map ...............................................................................
. Miamisburg, Ohio ......................................................................................
. Dexter Village, Michigan ...........................................................................
. Comparison of TOTAL and PT Slope/Intercept Values ..............................

. Comparison of TOTAL and LENGTH Slope/Intercept Values ...................

viii

page
25

29

3O

54

57

Chapter I:

Introduction and Background

Concern over human-made environmental hazards has increased within the last few
decades (Slovic 1986). Media attention regarding spills, accidents, and chemical
catastrophes has initiated heightened environmental attention within the United States.
Organizations and pro-active groups have formed to make citizens aware of current
Operations and technological advances impacting the environment. Within this larger
environmental context, perceptions concerning hazardous material transportation is a tOpic
that merits investigation.

Hazardous materials are identified as any material with one or more of the
following characteristics: ignitability, corrosivity, reactivity, or toxicity. Current estimates
of the amount of hazardous materials generated annually are 22 billion pounds
(Cunningham 1994). The United States Environmental Protection Agency (U .S.E.P.A.)
lists hazardous materials disposal, at both active and abandoned sites, as the first and
second most important concerns facing the American public (Roberts 1990). There is
recognition of hazardous materials in manufacturing and recognition of the disposal risks,
but little research has addressed the intermediary step of transportation of hazardous

materials.

2

Therefore, I would like to study the perceptions of two sets of respondents
concerning the rail transportation of hazardous materials. The study areas for this thesis
are Dexter Village, Michigan, and Miamisburg, Ohio. The study areas were chosen for
their location within the industrial belt, the amount of hazardous materials moving within
their state, and the towns' proximity to the rail lines.

The explicit study of environmental perception, as a theme of behavioral
geography and environmental psychology, began in the 19503. Besides focusing on one’s
" images of place," the early environmental perception literature concentrated on detailing
natural hazard perception (Burton and Kates 1964; White 1960; Calef 1950; Flora 1956).
This natural hazard perception concurred with the environmental determinists’ study of
environmental influence, or how the environment was thought to influence an individual‘s
perceptions and attitudes of the real world. Environmental perception literature reflected
upon the subjective images of the inhabitants to the hazardous areas (Arbuthnot 1977;
Borden and Francis 1978; Borden and Schettino 1979; Bruvold 1973; Calef 1950; Craik
and lube 1976; Dunlap and Van Liere 1978; Goodey 1974; Slovic 1986).

Within environmental perception research, the area of perceived risks has received
increasing attention. Slovic suggests that studies of risk perception should examine the
judgments people make when they are asked to characterize and evaluate hazardous
activities and technologies (1986). As geographers have attempted to explain
relationships between humans and the surrounding environment, studies of perception
have continued to develop. Past arguments have questioned whether one's activity within

the environment is determined not only directly by the form and characteristics of the

3

environment, but also indirectly by one's perception of the environment (Goodey 1974).
Today, risk perception research is concerned with providing a basis for understanding and
for planning how citizens will respond to a hazard and for improving the interaction of
technical experts within the field, the average citizen, and those in decision making
positions.

Research on risk perception has identified measurable factors that influence how
individuals perceive environmental issues. If important characteristics that predict
perception can be identified, it may be possible to aScertain how safe is safe enough.
Some of the individual characteristics that have been linked to environmental perceptions
include demographics, distance, and length of residence (Arbuthnot 1977; Borden and
Francis 1978; Borden and Schettino 1979; Craik and lube 1976; Dunlap and Van Liere
1978). Past research has been successful at identifying relationships between these
characteristics and perception. However, there is disagreement over which characteristics
are the best indicators of perception. This thesis will attempt to look at several
independent variables and to statistically determine if they are indicative of the perceived
risk fi'om hazardous material transportation.

Past studies have tried to describe risk perceptions in conjunction with
socioeconomic characteristics of the respondents. Questions regarding whether gender
and educational status have influenced the level of perception have suggested differing
accounts (Johnson and Buttel 1977; Slovic 1991). Geographic measures such as the
proximity to an industry or transportation corridor have also been discussed in perception

studies (Slovic 1994.) This thesis will examine several factors that may account for

4

variations in the perception of risk fi'om transporting hazardous materials. The factors,
previously discussed in the literature, which are incorporated in this study include gender
(Arbuthnot 1977), education (Bruvold 1973), length of residence (Lindell and Earle
1994), and proximity to the rail (Slovic 1991). One additional variable analyzed in this
study is whether a difference in risk exists between the two case study areas. This
comparison is incorporated in this study to ascertain what difl‘erences may exist between
an area where a rail incident has occurred and an area unafi‘ected by the transportation of
hazardous materials.

By examining socioeconomic factors which may influence risk perceptions, a better
understanding of the public's risk perceptions may be reached. Past literature has
analyzed what factors may be influential in determining risk levels (Bunting and Guelke
1979; Buttel and Johnson 1977); however, this study implements these factors into a
comparison of the two study areas. The first factor presented in this study is the role of
gender in determining perceived risk. Past risk studies involving gender have suggested
that women have a heightened sense of environmental awareness (Schahn and Holzer
1990) resulting in higher perceived risk levels. Other studies simply state that the
environmental and Equal Rights movements have coincided; therefore, leaving another
avenue for female empowerment (Borden and Francis 1978). This study will attempt to
contribute to the volume of risk perception literature by determining if gender is influential
in environmental risk perception.

Another socioeconomic factor which may contribute to perceived risk is the extent

Of one's education. Education has been touted as a key to environmental awareness.

5

Presumably, the higher the level Of one's education the greater the exposure to various
ideas (Arbuthnot 1977). Environmental activeness and awareness are thought to indicate
higher perceived levels of risk.

Geographic assessment is also necessary when trying to analyze risk perceptions.
The length of one‘s residence in an area and their proximity to the studied area (for the
purpose of this study: a rail line) have shown differing accounts in resident's perceived
risk levels. A lower level of perceived risk may be attributed to the greater length of
residence and the continued transportation of the materials without any exposure or
dispersal problems. Slovic M (1991) and Lindell and Earle (1994) concurred in their
studies that those living closer to the rail viewed the transporting of materials as safer.

The generation of hazardous materials and the treatment, storage, or
disposal of such substances has been the primary focus of past studies, while concern for
the risk of transporting the materials has received little attention. The transportation of
such materials allows for several factors to contribute to the overall risk. These risk
factors include any possible damages to the highway or rail track, weather conditions, or
routing concerns which may occur along the transportation route, ultimately resulting in
the increased likelihood of a dispersal.

With the strengthening of regulations concerning hazardous material disposal,
many generators of hazardous materials have found that treating or storing the material
onsite has become more burdensome than shipping the material to a treatment or disposal

facility. This burden has resulted in an increase in the transportation of materials to

Off-site facilities where the treatment process can continue. Rail transportation, the

6

subject of this thesis, has been one means of transportation used to transfer hazardous
materials from the production plants to the final treatment facilities. One reason for
utilizing rail for transporting hazardous materials is the fact that it is more than four times
safer than highway transportation. In terms of price, rail is cheaper than truck transport
due to the pricing scheme based on volume rather than distance hauled. Although motor
trucks are the primary mode of transporting hazardous materials, rail is used with great
frequency and safety- 1,497,464 origins in 1993, with 99.99% reaching their destinations
safely (American Railroad Association Annual Report 1993). While many governmental
entities have focused on highway transportation in their routing restrictions in urban areas,
few routing precautions have been incorporated in rail legislation

According to the American Railroad Association (ARA), there are two types Of
train mishaps: accident or incident. An accident is a situation in which a rail car of
hazardous materials derails, catches fire, explodes, suffers a serious sideswipe, or involves
death. It does not have to release its contents (but may), meet any dollar damage figure,
or any other requirement. An incident is the unintentional release of a hazardous material
while in transportation (including loading and unloading). This release does not usually
involve an accident (ARA Annual Report 1993). The focus Of this study is how citizens
viewed the risks of transporting hazardous materials by rail and what, if any,
socioeconomic aspects may contribute to the degree of risk.

Several risk factors for transporting hazardous materials have been identified. The
type of load, weather, rail condition, and time of day are potential contributors to any type

of rail mishap (Schierbl 1993; Saccomanno and Shortreed 1994). One problem addressed

7

by the ARA is that of material container types. While hazardous materials are transported
by rail in various types of rail cars, such as tank cars, covered hoppers, gondolas, box cars,
trailers, and/or containers on flat cars, the majority (70%) of hazardous materials move in
tank cars. In 1993, the number of train accidents in which a hazardous material car was
damaged or derailed was 262 (ARA Annual Report 1993). Of the non-accident releases,
89% were from tank cars.

Hazardous materials have long been transported by rail, yet little attention has been
given to the risk perceived by adjacent residents. As the role of transporting hazardous
materials by rail will only increase due to the continued production of such substances, the
need for proper assessment Of those risks perceived by the adjacent public is vital. Due to
the high volume of hazardous material generation and transportation within the industrial
belt of the United States, I have chosen to address risk perceptions in two towns within
this belt.

My working hypotheses are:

1. The residents of Miamisburg, Ohio, will indicate a higher total degree Of risk
perception concerning transportable hazardous materials than their Dexter Village,

Michigan, counterparts.

The higher perception of risk is expected due to Miamisburg's past experience with
a derailment, evacuation, and chemical dispersal in 1986. Another factor contributing to

Miamisburg's heightened awareness is its continued exposure to the DOE Mound facility,

8

which is currently undergoing corrective action. By first addressing comparisons between
the two communities I can further delve into variables which may be indicators of possible

reasons for certain perceptions.

2. Female respondents will indicate higher levels of perceived risk than the male

respondents.

The difl‘erence in gender perception is expected because of the need for women to
assert leadership within a societal movement or because of heightened environmental
concern. The focus in the existing literature has concerned the difference in how the sexes
perceive risk and why females are likely to indicate higher levels of risk. This study will

focus on the differences between the combined females and males in the two study areas.

3. Education is a significant indicator of perceived risk level, and those with

higher levels will perceive greater risk than those with less formal education.

The extent of one‘s education may relate to awareness of environmental issues and
current technologies in existence. Past literature has suggested a relationship between
education and perceived risk levels of environmental issues (Arbuthnot 1977). This study
attempts to test this education relationship in a comparative analysis of Dexter Village and

Miamisburg respondents' risk levels.

9

4. Proximity to the rail line is a significant indicator of risk and respondents living

closest will indicate higher risk levels than those at a distance.

Prior to addressing the relationship between the proximity of the respondent and
the rail line to the level of perceived risk, one must look at what distance defines a risk
exposure area. The establishment of such an area should be completed prior to the
distribution of a survey instrument. Slovic surveyed respondents in an exposure area of
two miles fiom the Oregon highway where hazardous materials were being transported
(1991). Within this two mile area Slovic found varying attitudes toward the transportation
Of hazardous materials. High risk perception levels were found for respondents living
nearer the transportation corridor within the two mile survey area. The two mile area
defined the distance at which a highway accident would disrupt residents, but the closer
their proximity to the rail line the higher their perception of risk. A two mile exposure
area was chosen as an adequate distance for the level of toxicity of the materials being
transported. When studying risk exposure areas, the extent of the area is based upon the
type and volume of chemicals being transported. Like Slovic, this thesis looked at risk
perceptions of residents up to two miles from the rail lines. Within both study areas, this
two mile zone encompasses the greatest concentration of the residents. Beyond this
distance, the population density decreases.

By defining my survey area I could assess the attitudes of residents at varying
distances from the rail lines. A greater concern is expected to exist for those adjacent to

the rail rather than those further away, because of the constant threat of not only a

10

dispersal but also a derailment or other accident (Flynn eLal 1990). Due to Miamisburg's
derailment history it is expected that the respondents closer to the rail were more affected
than those firrther away ultimately resulting in a greater risk perception. This study will
focus on the proximity of the respondents and analyze whether a respondent's distance

affects risk perceptions and whether the effects of distance are stronger in Miamisburg

than in Dexter Village.

5. The length of residence is a significant indicator of risk levels and respondents
with greater lengths of residence will indicate higher levels Of risk than those who

have resided in the areas shorter amounts of time.

Miamisburg respondents who have lived in the area prior to the 1986
derailment are likely to perceive higher levels of risk due to the derailment, evacuation,
and chemical exposure. Miamisburg's familiarity with the extensive problems of the DOE

Mound facility can contribute to their overall awareness.

Structure of thiiILapg

The general nature of the problem(s) and the risks associated with the
transportation of hazardous materials by rail, as well as the study's objectives and
hypotheses are presented in Chapter 1. Chapter II includes a detailed review of risk
perception literature pertaining to the hazardous material process and transportation,
variables which may influence the perceptions, and laws and regulations developed to

minimize the risks.

11

The study's methods, fi'om questionnaire development and distribution to study
area demographics are presented in Chapter III. Chapter IV presents the survey results
and statistical analysis in graphic, tabular, and map form. A comprehensive discussion Of
the study's results and their relationship to the primary hypotheses of this study is

presented in Chapter V.

Chapter II:

Relevant Literature

A discussion of the socioeconomic factors, their relation to risk perception, and

how they are integrated into this study will be presented in this chapter.

The New Environmental Paradigm

Considerable attention has been given to the emerging environmental
consciousness within the American society since the early 1970‘s. The advent Of
organizations, regulations, and grassroots activism has contributed to this overall interest
in natural resources. In a 197 8 article, Dunlap and Van Liere suggested that a “New
Environmental Paradigm“ (NEP) was emerging in society that confronted the older view
of an anthropocentric, anti-ecological order. By constructing a 12-item NEP scale, with
Likert-formatted questions, the opposing views of how humans regard the natural
environment was tested. After subjecting the data to analysis, Dunlap and Van Liere
concluded that the 12 items which comprised the NEP scale formed "an internally
consistent and unidirnensional scale (1978)." Dunlap and Van Liere's analysis is a
prototype for risk perception studies such as this comparative assessment of two towns
concerning hazardous materials transportation. The NEP scale is necessary in developing

an instrument by which to assess perceptions. Dunlap and Van Liere’s new environmental

12

13

paradigm describes this emerging concern for one's own health and that of the natural
environment. The NEP is integral to perception studies as it reflects the growing trend

toward environmental sensitivity.

Risk E erts vs. La le

Risk assessment is a decision making tool in environmental planning. Risk
assessment is viewed as one part of risk analysis by the Council on Environmental Quality
(Cohrssen and Covello 1989). It is the second step in a process which includes identifying
the hazard, assessing the risk, determining the significance of the risk, and communicating
the risk. The assessment phase estimates "the severity and likelihood of harm to human
health or the environment“ (Cohrssen and Covello 1989). It is often used in controversial
planning situations as the basis for regulatory decisions including standard setting and
permit approval. Risk assessment can heighten conflict in some planning situations
because it provides numbers to debate.

The result of risk assessment is a probability of harm from exposure to the
materials and processes. The probability is used to identify management strategies to
reduce the risk to de minimus levels. De minimus non carat [ex is a legal doctrine that
means the "law does not care for, or take notice of, very small or trifling matters" (Black
1990). De rninimus risk is referred to as a risk of one in a million during a 70 year
lifetime. Risks are subjective though, as some people view an estimate of deaths from
cancer (ex. 1 in 1,000,000) as too much, while others dismiss the risk. The disagreement

is not over the estimated numbers, but in how peOple perceive the numbers. Even when

14

the source and accuracy of risk assessment data are agreed on, there is Often conflict
concerning the differences in perception of the data.

While the subjectiveness of risks are debated, a discrepancy also exists between
these experts active within the environmental field and the average citizen with little
technical knowledge. Concerning the perceived risks associated with hazardous materials,
the technologically-competent analysts are inclined to rely on risk assessments, while the
vast number of citizens use intuitive risk judgments or risk perceptions. One of the best
examples of this contrast is evident in the United States Environmental Protection
Agency's (U SEPA) survey comparing the American public's environmental concerns with
those of the USEPA The American public listed active and abandoned hazardous waste
sites as the numbers one and two highest concerns (Roberts 1990). See Table 1. Of the
top ten environmental risks, an additional four were concerned with the exposure to
various chemicals as a result of accidents or releases from industrial sites. The heightened
concern for hazardous material handling demonstrated in the USEPA survey provides the
basis for surveys such as mine. The insecurities exhibited by the USEPA's public survey
respondents concerning hazardous materials on contained sites is the foundation for
examining the attitudes toward the same materials in a moving environment. In contrast to
the public’s risk perceptions, the USEPA's highest concerns did not mention hazardous
waste sites.

Sandman, egg (1993), suggest that the public-versus-expert risk controversies are
due to the public's lack of understanding concerning scientific data. This " science

illiteracy" is addressed by using venues like the mass media to distribute technical

15

information. Starr (1985) suggests that the "acceptance of any risk is more dependent on
the public’s confidence in risk management than on the quantitative estimates of risk. "

This thesis looks at the perception of risks of the common citizen. By choosing to
study two areas, I was able to examine the perceptions of two sets of citizens who have
had varying degrees of interaction with hazardous materials. The extent of this variance

will be examined in Chapter IV.

Factors Influencing Risk Permtion

Risk perception research is concerned with providing a basis for understanding and
planning how citizens will respond to various scenarios. An integral part of assessing risk
is determining if any background factors contribute to the level of perceived risk. Several
suggestions have been given concerning what demographic and socioeconomic factors
may contribute to environmental interest. Four variables (education, gender, length of
residence, and proximity to the rail) selected from the previous literature concerning risk
perception and introduced in Chapter I, will be incorporated into the analysis of the
responses from Dexter Village and Miamisburg for further comparison of risk perceptions
between the two study areas. Each variable will be discussed in this chapter and then

analyzed within Chapter IV.

16

Table l: USEPA Survey of the Public's Perception of Environmental Risks

lg'sk Percent of Respondents who Indicated VERY SERIOUS
Active Hazardous Waste Sites 62
Abandoned Hazardous Waste Sites 61
Worker Exposure to Toxic Substances 60
Industrial Water Pollution 58
Nuclear Accident Radiation 58
Radioactive Waste 55
Underground Tank Leaks 54
Pesticide Harm to Users 54
Pesticide Residues 52
Industrial Accident Pollution 51
Farm Run-off Water Pollution 50
Tap Water Contamination 49
Industrial Air Pollution 48
Ozone Layer Destruction 47
Sewage Plant Water Pollution 45
Vehicle Exhaust 38
Tanker, Rig Oil Spills 38
Acid Rain 36
Urban Run-Ofl' Water Pollution 35
Damaged Wetlands 35
Genetic Alteration Damage 35
Nonhazardous Waste Sites 33
Greenhouse Effect 33
Indoor Air Pollution 26
X-Ray Radiation 22

Source: Richard Morgenstem and Stuart Sessions, "Weighing Environmental
Risks: EPA'S Unfinished Business" Environment 30 (1988): 14-39.

l7

Constantini and Hanf submit that an individual's degree of concern may be closely
associated with more firndamental social and political attitudes as well as general social
role and background (1972). One of the dominant characterizations is that environmental
consciousness has been largely a middle and upper class phenomenon, which is measured
in terms of education, income, and occupation (McEvoy 1971; Hany M 1969). Harry
gt_al discussed the attitudes of those involved in conservation associations in the Pacific
Northwest and found that membership in the movement appears to be composed largely of
upper-middle-class citizens (1969).

Environmental concern may also be attributed to education and past experience.
In their study of Lake Tahoe residents, Constantini and Hanf described those citizens who
were better educated as displaying a higher level of environmental concern (1972).
Arbuthnot concurred with this sentiment that, in the Athens, Ohio, area, those persons
with higher educations practiced a higher degree of recycling and had more environmental
concerns in the Athens, Ohio area (197 7). Presumably, those citizens with greater levels
have been exposed to more ideas and are more apt to be involved within the community.
This study analyzes the educational levels of the respondents and statistically compares the
relationship to perceived risk in the two study areas.

Gender is an additional factor that may contribute to how one perceives a risk.
Past studies have suggested differing accounts, as positive relations in one study (Van
Liere and Dunlap 1981) are reversed in another (Arcury, Scollay, and Johnson 1987).
Increased environmental concern by women has been reported as higher values in attitude

scales and measures of self-reported behavior, but less knowledge about environmental

18
problems has been displayed by women (Schahn and Holzer 1990). Borden and Francis
suggest that the increased environmental concern by women may be attributed to a
potential for leadership, as the environmental movement has evolved with the women's
movement (197 8). Slovic etg discuss that female respondents indicated the transportation
of hazardous materials as less safe than did their male counterparts (1991).

Slovic 1a; described in their study of transporting hazardous materials by highway
the demographic aspects of the respondents along the corridor (1991); those respondents
living closer to the corridor saw transportation as safer. Lindell and Earle concurred that
the perception Of risk from hazardous facilities is often lower for those living closest to the
facilities (1994). Slovic eLa_l_ analyzed the responses of Oregon residents up to two miles
from the highway. The two mile distance fi'om the highway is considered the risk
exposure area or the zone in which an accident involving the transported materials could
put the adjacent community at risk for evacuation or potential health problems. Slovic's
study looked at a relatively small-scale exposure area due to the volume and type of
materials being transported. Other risk exposure areas can extend several hundred miles
from the site of the accident. Recent events resulting in enormous higher order impacts
include the chemical manufacturing accident at Bhopal, India, and the meltdown of the
nuclear reactor at Chemoybl. Following these extreme events are a myriad of mishaps
varying in the breadth and size of their impacts, ultimately affecting larger populations
within greater exposure areas (Saccomanno and Shortreed 1993). Planners have attempted
to decide at what distance does a risk dissipate (Saccomanno and Shortreed 1993). The

need to define a consistent distance is unnecessary as each dispersal is site-specific. Past

l9

literature has looked at small-scale case risk exposure areas such as those analyzed in this
study (Slovic 1991). The determination of a risk exposure area is dependent upon the
population density, type of material, and volume of the material. The density of the two
study areas combined with the proximity of the town populations to rail lines provides for
a representative sample to exist within the defined two mile area.

One's length of residence in the community can contribute to their perceived risk
level. Those who have been subjected to continual exposure without any difficulty may be
more likely to indicate a lower level of risk. A lower level of risk may be attributed to the
greater length Of residence due to the continued transportation of the materials without
any exposure or dispersal problems.

As hazardous material usage continues to increase, the issue of treatment of the
material has become an increasingly complicated issue. As the difiiculty in treating
materials and waste on-site has escalated, the importance of reliable transportation to such
commercial outlets has been growing. This reliance upon transportation, especially rail, as
one component of the hazardous material cycle is the primary reason for further exploring

the public’s attitudes and reactions to the corresponding risks involved.

P i n n 'n Tr in H ri

Due to the increase of regulations regarding hazardous materials, many producers
of hazardous wastes and materials have resorted to Off-site, licensed disposal facilities and
storage units. This increase in off-site treatment has resulted in increased transportation of

the hazardous materials. Analysis by the United States Department of Transportation

20
indicates that hazardous material shipments impose significant risks to the population.
Between 1971 and 1980, more than 111,000 accidents involving hazardous materials were
reported, resulting in a total of 248 fatalities, 6873 injuries, and approximately $120
million in property damage. Railroad accidents accounted for 8% Of the accidents and
18% of fatalities (National Transportation Safety Board 1981).

Few researchers have described the risk perception of transporting any type of
hazardous material. One of the first studies was done by Slovic et_a_l in Oregon. It was an
initial description of a program to transport radioactive waste fi'om the Hanford Reserve in
Washington through Oregon by highway (1991). Of the Oregon residents surveyed,
69.1% agreed that present methods for transporting hazardous materials out of the
Reserve are necessary for proper remediation activities to occur. Regarding current
methods for transporting hazardous materials, half of the respondents (50.7%) concluded
that today’s techniques are reasonably safe; however, respondents viewed transportation as
a riskier process than storage.

Flynn :13! studied the perceptions of risk and benefits associated with a nuclear
waste repository, trust in the Department of Energy, and views on other issues pertaining
to nuclear waste (1990). Their study indicated that people perceived the transportation Of
nuclear waste as highly risky. The percentage of respondents who agreed or strongly
agreed to “Highway and rail accidents will occur in transporting the wastes to the
repository site, " was 77.4% in Nevada, 69.2% in California, and 71.6% nationally (1990).

The process of transporting hazardous materials is integral in the reduction of the

volume of materials being stored on site in a particular region or community. This

21

reduction in the volume of materials stored on site may be a benefit to the community
shipping the materials, yet it may instantly become a nuisance for the receiving community.
The answer to this dilemma is found in acceptable, safe procedures used to permanently
deter the materials from infringing upon the natural environment or the surrounding
communities. Safe procedures are vital, but the reality of transporting hazardous materials
is evident within the two study areas. The focus of this study is on the perceived risk of
the transportation of such materials and what factors may influence the degree of risk
perceived.

Beyond assessing environmental risks, Oregon respondents concurred with many
of the images offered by those who are reactive to residential facility sitings. Respondents
believed that areas through which radioactive wastes are transported are likely to be
unattractive to business development and tourism. Respondents also indicated little
confidence in the state of Oregon to provide accurate information concerning the
transportation process. Within the survey, 88% of the images evoked by word association
were negative (Slovic 1991).

The transporting of hazardous materials results in several issues which can benefit
from perception studies. The ”unattractiveness" of areas where routing may occur can
lead to undevelopment and lower land values. The lack of information flow fi'om those in
authoritative positions can Often result in little confidence in ofiicials dealing with the
transportation process. The risks to the health of the environment and the citizens who

are subjected to the materials through routing plans can be better assessed.

22
Regulations

Most of the environmental regulations concerning the transportation of hazardous
materials have been formulated in response to the public's perception of various levels of
risk. Federal regulations (49 CFR Part 397) have established standards for maintaining
and enforcing designated highway routes for truck shipments of hazardous materials.

Ohio's regulations were reactionary due to the derailment of the train in
Miamisburg. Ohio House Bill 428 was signed on June 24, 1988, in Miamisburg by then
governor Richard F. Celeste in response to the derailment which had occurred two years
earlier. With the implementation of Ohio's Hazardous Material Transportation Act
(OHMTA), a registration program for all haulers was created along with a fee collection
and disbursement program for emergency response training. The OHMTA also instituted
a new system of enforcement by imposing civil forfeitures rather than criminal penalties.
In relation to rail transportation, rules in Ohio and Michigan are consistent with the
Federal regulations. Both states invoke regulations similar to 49 CFR.

The relevance of Ohio House Bill 428 to my study is important, as the very people
who were affected by the derailment are those same individuals questioned by this survey.
The events which happened in Miamisburg brought awareness of the risks of transporting
materials through communities. Prior to this event, few regulations were in place. The
effectiveness of these regulations in difi’using any potential risks to the citizens may be
established by analyzing their perceptions of the continuing transportation of such
materials. As mentioned previously, perception of risk is a vital prerequisite of the

planning process and creation of legislation.

Chapter 11]:

Methods

This chapter discusses the procedures I used to numerically rank and
geographically analyze the risks associated with general environmental issues and
transportation risk factors. Divided into four parts, this chapter focuses on research
design, including the Objectives for the study. Sections two and three of the chapter
concern the study area and the survey instrument respectively. The final section includes

the techniques for questionnaire distribution and analysis of the respondent demographics.

Research Desigg

This analysis was designed to contribute to the understanding of risk perception
and to describe the attitudes of residents concerning the transportation of hazardous
materials. Utilizing proximity, socioeconomic variables, and a case study approach, this
study answers the question of how the two study areas compare and contrast in their
views on transportable hazardous materials by rail. The proximity analysis is useful in
establishing relationships that may exist between location and degree of perceived risk. In
addressing the risk perceptions of hazardous material transportation, I will address how
independent variables of gender, education, town, and length of residence contribute to

awareness (Borden and Francis 197 8; Schahn and Holzer 1990).

23

24

This research is consistent with Slovic‘s body of research, in that it attempts to
measure the perceptions of risk in transporting hazardous materials and subject the results
to statistical analysis. Slovic has discussed the use of the "psychometric paradigm" which
has explored the usefirlness of psychophysical scaling methods combined with those of
multivariate analysis in producing representations Of risk attitudes and perceptions (1986).
Past researchers have used the psychometric paradigm in asking people to judge the
riskiness or safety of various hazardous activities, materials, and technologies and to
indicate their desires for risk reduction and regulation of these hazards. These judgments
have then been related to decisions about the hazards position on various qualitative

characteristics of risk.

Study Area

To better explain attitudes toward the transportation of hazardous materials by
rail, I have analyzed and compared the responses of residents in two different areas, one
of which has experienced an accident and one of which has not. The two areas, Dexter
Village, Michigan and Miamisburg, Ohio were chosen for the comparable socioeconomic
composition of their populations, railroad proximity to the population, population size,
and the fact that hazardous materials are being transported on the railroads through these
areas. Michigan and Ohio's location within the industrial belt of the United States and

their past records of train incidents were also considered.

25

Figure 1: Dexter Village, Michigan and Miamisburg, Ohio

CASE STUDY AREAS

DEXTER VILLAGE, MICHIGAN

 

MIAMISBURG, OHIO

 

26

Miamisburg, Ohio

Miamisburg, Ohio, situated in the southwest corner of the state, is located 10 miles
southwest of Dayton. The United States Census lists the population of Miamisburg as
17,834 (1990). The railroad tracks are situated between houses through the center of the
city. The income and educational levels of Miamisburg residents are diverse while the
racial composition is homogeneous (98% Caucasian).

Miamisburg was also the scene of the largest evacuation in the United States due
to a train accident and chemical dispersal. On July 8, 1986, a southbound Baltimore and
Ohio train traveling from Walbridge, Ohio, to Cincinnati derailed in Mianrisburg. The
train was carrying thirty-five cars, several of which were chemical containers. While
traveling across the Great Miami River, at a speed of 45 mph, the train hit an area of track
which had expanded due to repair work and the heat, causing the 24th thru the 38th cars
to derail. The 30th car, a load of white phosphorous, ignited and burned, resulting in the
evacuation of a two mile area adjacent to the rail line (PUCO Report 1986). Area fire
departments arrived immediately and began to remediate the bunting phosphorous by
applying water. This process continued for 24 hours (into July 9th) at which time the fire
stopped and water spraying ended. At 6:05pm on July 9th, the phosphorous car reignited,
and exploded, resulting in a second evacuation. According to Bill Mitchell, a Federal
Railroad Inspector at the scene, the phosphorous tank was sprayed with water again until
it was finally decided on Friday July 11th to vent the car and install fans for a hot and fast
burn. By 7:00 am, Saturday July 12th, the fire was finally extinguished (Bill Mitchell,

personal interview, July 1995).

27

Miamisburg is an ideal location for assessing how people perceive hazardous
material transportation risks, as the 1986 deraihnent was the largest of its kind. The
evacuation of over 15,000 residents made startlingly evident the idea of hazardous
material transportation risks. Miamisburg is not only a favorable area due to the
derailment, but situated at the southern tip of the city is the United States Department Of
Energy Mound Facility. DOE Mound manufactures non-nuclear and tritium-containing
weapon components. In their production processes, Sanitary, hazardous, radioactive, and
mixed wastes are generated. Environmental restoration currently includes a clean up
program with decontamination and decommissioning activities. Miamisburg residents'
interaction with hazardous materials is a daily issue and it is for this reason that I wanted
to try to better assess their perceptions Of risk.

The initial derailment of the phosphorous tanker resulted in the evacuation of the
respondents within a half mile of the spill. The risk exposure area of the Miamisburg
derailment was broadened due to the actions of the emergency response crews. The
phosphorous fire that resulted fi'om the derailment was doused with water causing a
mushroom cloud to form and move down through the Miami Valley. This resulted in a

larger exposure area and a total of 15,000 people being evacuated.

Dexter Village, Michigan

Dexter Village, Michigan, situated in the southern lower peninsula, is located
approximately 11 miles northwest of Ann Arbor. The village population is listed as 1,497

with most residents residing within one mile of the railroad tracks. Income of the

28

residents of Dexter Village are highly varied. The residents are racially homogeneous
(Caucasian), but vary in educational levels and occupations (1990 United States Census).
Dexter Village does not have a history of rail accidents, but the line that runs through the
town is actively transporting hazardous materials.

Miamisburg and Dexter Village are small towns, yet physically adjacent to small
cities (Dayton and Ann Arbor). This proximity facilitates emergency planning, by the
LEPC (Local Emergency Planning Committee) and it provides a larger body of
contingency workers capable of handling emergencies involving hazardous materials. One
interest of mine was to see if information about the transporting of hazardous materials
had been dispersed to the residents adjacent to the route. A respondent's higher perceived
level of risk could be attributed to this information flow.

This research was designed to analyze geographically the perceptions of risk in
transporting hazardous materials through two study areas by rail. By partitioning regions
within the study areas to distribute questionnaires and trace responses, I could subject all
responses to proximity measurements. For each of the study areas, I analyzed aerial
photographs of the regions and then partitioned the 20 residential blocks closest to the rail

line as my study regions.

29

Figure 2: Miamisburg, Ohio

 

Scale: 1224,000
Source: United States Geological Survey. State of Ohio, Miamisburg Quadrangle. 1987.
(Enlarged 10%)

30

Figure 3: Dexter Village, Michigan

    
    

53"., V "-
7," ‘ \ .. 37 “(a 3:. "All ’3
' o 1. _~-\ \ - ,/' Y k‘ I
' ..., fl, -\ \. , var/"x \u 4.3: y t l
. I I ,, a; . -
’ L2“ . ,
' 'm , . . '
“L3
‘ I
l ' 1" DAN awas-

Scale: 1:24,000
Source: United States Geological Survey. State of Michigan, Dexter Village Quadrangle.
1987. (Enlarged 10%)

31

urv Instrument

Risk perception is the threat that an individual believes is associated with a specific
environmental issue. The measurement of risk perception in this study is accomplished by
assessing opinions. Three categories of risk perception are used: the perception of
environmental risks in general, the perception of risks associated with the transportation of
hazardous materials, and risks associated with specific factors in Dexter Village and
Miamisburg (such as the Huron and Great Miami Rivers). The survey instrument utilized
to assess the risk perception was a questionnaire with closed-ended questions (see
Appendix C). The questionnaire was divided into three sections corresponding to the three
categories above. A fourth section detailing respondent demographics was also included.
Sections one through three allowed the respondent to indicate their risk perception level
on a Likert style rating scale of 0 (low risk) to 6 (high risk) (de Vaus 1990). Section four
allowed for both fill-in style questions, scale, and yes or no responses. The questionnaire,
which was gee—coded and stratified by zone begins with general questions concerning
human health and the environment and then progressively becomes more site-specific.

The first section of questions required the respondents to indicate the level of risk
they associated with various environmental pollutants. The environmental issues for the
general perception measure came from a comparative risk assessment conducted by the
United States Environmental Protection Agency. In that study, EPA scientists defined and
ranked 31 environmental issues on the basis of human health, ecological, and welfare risks
(Morgenstem and Session 1988). For this research only 8 of those issues were included in

the questionnaire, including the threat of water contamination and the possible exposure of

32

the rivers within the two towns. Worker exposure is an inherent risk factor in any type Of
hazardous material transportation as is a decrease in air quality should a dispersal occur.

The second section of the questionnaire looks at factors of transporting hazardous
materials and the extent to which they are seen as a risk to the environment. The
transportation risk factors used in this study were delineated by the Oflice of Technology
Assessment in their report on the Transportation of Hazardous Materials (1988). These
factors are concerned with the (1) condition of the rail, (2) weather factors, (3) type of
load, and (4) amount of material being transported. Saccomanno and Shortreed
acknowledge these factors in their study of the perception of societal and individual risks
associated with transportable hazardous materials (1994). Again, respondents indicated
their perceived risk on a Likert response scale from 0 (NO Risk) to 6 (High Risk).

The third section presents environmental features which may be at risk within the
study areas. Both study areas have rivers adjacent to the towns. Other potential risks
included those associated with the proximity to the tracks, the vulnerability of the towns,
and the respondent's home.

The final section of the questionnaire asked the respondent to indicate various
socioeconomic attributes. The demographic factors were included for further analysis of
any correlations between the five independent variables (gender, education, income,
proximity, and length of residence) and level of risk perceived. Following the
demographic factors were either three (for Dexter Village) or four (for Miamisburg)
additional site specific questions. The survey for Dexter Village included an inquiry into

whether the respondent was aware of any transportation of hazardous materials through

33

the village. The awareness of the residents was questioned due to little acknowledgment
by the rail and county oflicials of rail transportation of such materials in Dexter Village.
The main focus of the Washtenaw County officials has been on the highway routing Of
such materials. The July 1986 derailment and evacuation in Miamisburg made such a
question unnecessary there. Other questions asked if the respondent was involved in any
type of environmental Organizations and if they felt that their elected Oficials were
addressing environmental issues.

The survey instrument allows the respondents to indicate their perceptions of risk
for general environmental issues as well as factors concerned specifically with the
transportation of hazardous materials. Since each issue had a scale fi'om 0 to 6, three
intermediary scores (environmental issues, transportation issues, and site issues) and a

final score for the total level of risk can be tallied for each respondent questionnaire.

lira-lest

According to Dillman's (1978) outline for pre-testing, the questionnaire was
administered to three sets of people: professionals within the field, others associated with
the field, and others outside of the field who represent a cross section of the population to
be studied. For the purpose of this study, I distributed 10 pre-test questionnaires. Four
respondents were employees of the Ohio Environmental Protection Agency, Division of
Hazardous Waste Management. Three test respondents were fiom the Ohio
Environmental Protection Agency employed in positions that are not directly related to

hazardous waste management but support divisions such as Data and Systems and

34
Groundwater. The remaining three respondents came fiom outside of the environmental
field and were representative of the intended study respondents.
The feedback from the pro-test allowed for changes to the questionnaire prior to
the distribution of the main survey. The comments fi'om the pre-test group indicated a
need for clarity of the questions (re-wording) and a need for examples of how to read and
answer each statement. Once these changes were made, I resubmitted the questionnaire to

the pilot group for feedback. The clarifications addressed the pilot groups' concerns.

Survey Distribution

The questionnaire was distributed by a stratified scheme (de Vaus 1990) in both
Dexter Village and Miamisburg. This stratification process was developed fiom aerial
photographs of each area. A breakdown by city block was seen as the most optimal way
to gage proximity to the railroad and the easiest for recording responses. The Dexter
Village and Miamisburg study areas included the twenty blocks (zones) closest to the
tracks.

As mentioned in the previous chapters, the surveys were distributed at homes up to
two miles from the rail lines. This distance was chosen as the risk exposure area due to
the population density and layout of each town relative to the rail line. The two mile
radius also encompassed the evacuation area in Miamisburg from the 1986 derailment.
Beyond the two mile zone, the population density decreases.

The 100 Dexter questionnaires were distributed on July 29, 1995 and the 100

Miamisburg surveys were delivered on August 19, 1995. Both questionnaires were

35

delivered by hand with stamped envelopes in the survey packet for responses to be
returned by mail (Dillrnan 1978). In a follow-up to the initial questionnaire distribution, I
targeted those zones from which I had not received sufficient initial survey responses and I
returned to Dexter Village on September 10, 1995, and distributed 25 additional surveys.
On September 21, 1995, I distributed 15 additional surveys in Miamisburg. In total 53

surveys (22%) Of the 240 distributed were returned completed.

Table 2: Socioeconomic Profile of Dexter Village and Miamisburg

Dexter Village, MI Miamisburg, OH
Population 1,497 17,834
Number of Households 676 6671
Median Household Income $32,411 $32,436
Racial Component 98% Caucasian 98% Caucasian
Educational Attainment ‘
(Person 25 yrs and older) 1,019 11,557
High School Grad. 293 (29%) 4317 (37%)
Some College 250 (25%) 2086 (18%)
Associate's Degree 48 (5%) 769 (7%)
Bachelor's Degree 156 (15%) 1246 (11%)
Graduate Degree 88 (9%) 336 (3%)

Source: 1990 United States Census

36

Analytical Methods

After receiving the responses back, I began to sort through the data and begin my
analysis. By dividing the questionnaire into three sections, I could use the Environmental,
Transportation, and Site issue scores individually or combined. The combined risk levels
were calculated by adding the first three Sections of the questionnaire together for each
respondent. I then completed preliminary analysis including the means and standard
deviations of the issues to ascertain what similarities and differences existed between the
two areas. The calculated means of the issues were usefirl in establishing a ranking of the
risks.

Once the introductory analysis was completed, I could then focus on the testing of
my hypotheses. My hypotheses were based on relationships between socioeconomic
variables and the perceived risk levels. The methods of statistical analysis I used were
t-tests, correlations, and regression analysis to establish what relationships did exist

(Slovic 1986).

Chapter IV:

Analysis

In this chapter, the data gathered fi'om the survey instrument outlined in Chapter
III are analyzed and the survey results are presented. I will examine the general
hypotheses that are the basis for the research, analyze the variables introduced in Chapter
I, and discuss the relationships among the variables.

The survey measured several variables including risk perception, gender,
education, and proximity to the rail. Table 2 summarizes these variables and defines the
categories associated with each. The risk perception values were measured on scales,
while the socioeconomic values were categorized on the survey instrument and indicated
by the respondent. The following discussion provides greater detail about each of the

variables.

Risk Perception

The respondents identified a level of risk represented by a total of 31 issues on a
7-point scale, with 0 being "NO Risk" and 6 being "High Risk." Tables 4 and 5 compare
the mean perception scores for each of the general and transportation issues presented in
the survey. Table 4 shows the perception of human health and environmental risk fiom
the general environmental issues, while Table 5 shows the perceived risk for the

transportation issues. Table 6 represents the mean perception scores for the site risk

37

38

factors. The general environmental risks have been identified by the Environmental
Protection Agency (EPA) as those issues dominating their agenda at all levels of

government.

Table 3: Summary of Variables Used in Analysis

ENVIRO: Total General Environmental Issues Score; Sum of the sixteen questions that

measure respondents perceived level Of risk to human health and environment.

TRANS: Total Transportation Risk Factors Score; Sum of the ten questions that measure

respondents perceived level of risk.

SITE: Total Site-Specific Issues Score; Sum of the five questions that measure

respondents perceived level of risk.

TOTAL: Total Risk Perception Score, Sum of the ENVIRO, TRANS, and SITE scores.

SEX: Gender of Respondent; Indicated by the respondent on the survey instrument.
Dummy Variable l = Female, 0 = Male

LENGTH: Length of Residence; Indicated by the respondent on the survey instrument;
categorized into 7 groups: Less than 1 year, 1 to 2 years, 2 to 5 years, 5 to 10 years, 10
to 15 years, 15 to 25 years, and Greater than 25 years.

EDUC: Highest Education Level Attained, Indicated by the respondent on the survey;
categorized into 6 groups: Some High School, High School or Equivalent, Some College,
College Completed, Technical School, or Graduate School.

39

Table 3 (cont'd)

REGION: Geo-coded Residential Blocks determined prior to the Survey Distribution.

TOWN: Study Area Designation, Dummy Variable, l = Miamisburg, 0 = Dexter.

PROXIMITY: Distance from the railroad, estimated in quarter-mile increments from the
center of the designated zone to the rail line.

ET: The EDUC level of the respondent multiplied by the respondent's TOWN.

PT: The PROXIMITY variable multiplied by the respondent's TOWN.

LT: The LENGTH variable multiplied by the respondent's TOWN

The mean perception scores were calculated for each section prior to testing the
hypothesis. Of the general environmental issues, Dexter Village and Miamisburg
respondents indicated 3A "Industrial Discharges to the Surface Water" as the highest
perceived level of risk to human health with the mean of 4.774 (sd =1.310). "Industrial
Discharges to the surface water" was the risk to the environment that had the highest

mean perceived risk level, 4.925 (sd= 1.378).

40

Table 4: Mean Perception Scores for the General Environmental Issues

(and Standard Deviation)

Perceived Perceived

Risk to Risk to
Risk Factor Rank Human Health Rank Environment
Air Pollutants 3 4.472 (1.370) 3 4.585 (1.262)
Toxic Air Pollutants 2 4.755 (1.385) 2 4.604 (1.378)
Industrial Discharges 1 4.774 (1.310) 1 4.925 (1.378)
Hazardous Waste 5 4.038 (1.480) 5 4.377 (1.484)
Accidental Releases 4 4.302 (1.381) 4 4.566 (1.487)
Radiation Exposure 7 3.811 (1.582) 6 3.736 (1.711)
Drinking Water 8 3.245 (1.440) 8 3.00 (1.732)
Worker Exposures 6 3.906 (1.510) 7 3.566 (1.575)

The second section of the questionnaire concerned the familiarity of the
respondents to issues concerning the transportation of hazardous materials. Within
hazardous materials management, there are several factors, identified in Chapter H that
can be threats to the environment. Risk factors were again rated on the Likert scale of 0
(Low Risk) to 6 (High Risk). The respondents rated number 7 "Condition of the rail and
tank cars" as the factor associated with the highest level of risk (4.811, sd=1.302.) I
expected the "condition of the rail" and "population density" to be the highest rated due to

Miamisburg's past derailment.

41

Table 5: Mean Perception Scores for the Transportation Risk Factors

(and Standard Deviations)
Risk Factor Rank Perceived Risk Level
Type of Materials 8 4.094 (1.522)
Volume of Materials 3 4.321 (1.451)
Emergency Response Capability 2 4.340 (1.518)
Travel Distance 4 4.302 (1.612)
Population Density 5 4.283 (1.486)
Natural Resource Density 7 4.245 (1.568)
Condition of the Rail 1 4.811 (1.302)
Time of Day 10 3.981 (1.704)
Weather Conditions 9 4.019 (1.647)
Type of Storage 6 4.264 (1.534)

The third section of the survey instrument, with risk factors that are site specific to
each town surveyed, yielded a highest rating for "the respondent's home," while the "town
population” was considered at the least risk in the process of transporting hazardous
materials.

I expected the "tank car derailment/accident" and "Dexter Village/Miamisburg
populations" would be perceived to be at the greatest risk. I submit that the past
derailment and the respondent's comments contributed to the above ranking.

Until now, I have focused on perception of risk fi'om individual issues. The
specific issues (ENVIRO, TRANS, and SITE) are useful in looking at specific risk
perception. However, it is interesting to examine the distribution of the total risk

perception score of all the issues. The total score is the sum of the risk perception

42

responses to all 31 issues and constitutes the individual's .overall perception of the issues.

The mean total risk perception score is 128.1 and the standard deviation is 36.11.

Table 6: Mean Perception Scores for the Site Risk Factors

(and Standard Deviations)
Risk Factor Rank Perceived Risk Level
The Huron/Great Miami River 2 4.377 (1.509)
Respondent's Home 1 4.698 (1.671)
Town Population 5 4.075 (1.627)
Chemical Dispersal within Town 4 4.170 (1.477)
Tank Car Derailrnent 3 4.264 (1.778)

Demographic Profile

The demographic profile of the respondents revealed several similarities between
the two study areas. The median ages Of the survey respondents is quite similar (45 and
48). While Dexter Village had more females than males respond, Miamisburg had a
greater number of males respond. The education and income levels were distributed fairly
evenly with the largest discrepancies in the SOME COLLEGE and GRADUATE
SCHOOL categories. The education and income levels were comparable to those
indicated by all Miamisburg and Dexter Village respondents in the 1990 United States
Census.

The racial composition of the respondents was 100% Caucasian in Dexter Village

and 96% in Miamisburg, which were figures comparable to the 1990 Census profiles.

43

Responses were returned fiom all forty geographic zones (20 in each study area). The
questions at the end of the Background section concerned one's involvement in
environmental activities, awareness of hazardous material transportation in their
community, and the efi‘ectiveness of elected officials. Twenty-two percent of the Dexter
Village respondents indicated involvement in an environmental organization, while
sixty-six percent of Miarrrisburg respondents stated involvement. In Dexter Village,
nineteen percent of the respondents stated that they were aware of Conrail transporting
hazardous materials and eleven percent suggested that elected Officials were “actively
pursuing ways to mitigate" transportation incidents. Seventy-one percent of the
Miamisburg respondents were living in Miamisburg at the time of the derailment, while
sixty-six percent were evacuated. Thirty-eight percent of the Miamisburg respondents felt
that the elected officials were "actively pursuing ways to rrritigate" transportation

incidents.

Demographic Issues

The issues of gender, education, length of residence, and proximity to the rail were
identified by the respondents in section IV of the survey instrument. These data were then

incorporated into the analysis (as the independent variables) for testing the hypothesized

44

Table 7: Demographic Profile of Survey Respondents

Dexter Village, MI Miamisburg, OH
Number of Responses 27 25
Age (Median) 45 48
Sex Female (65%) Female (44%)
Male (35%) Male (56%)
Education
Some High School 4% 8%
High School/GED 22% 16%
Some College 18% 40%
College Completed 26% 28%
Technical School 4% 4%
Graduate School 26% 4%
Length of Residence 6- 10 Years 6—10 Years
Race 100% Caucasian 96% Caucasian
4% Native American
Income
$0- 7,000 0% 4%
$7,000-15,000 7% 16%
$15,000-22,000 15% 12%
322,000-3 0,000 1 1% 24%
$30,000-40,000 15% 16%
$40,000-50,000 30% 8%
$50,000-60,000 7% 16%
$60,000 and greater 15% 4%

relationships. The gender, education, and length of residence were indicated on the
questionnaire, while the proximity to the rail was tracked through my initial geO-coding
process.
W2

Before proceeding with t-test and regression analysis to explore the hypothesized

relationships presented in the first chapter, I calculated bivariate correlation coemcients

between the independent variables (EDUC, PROXIMITY, and LENGTH). TOWN and

45

SEX were not included because they were Dummy variables. The purpose of calculating
correlation coefiicients is to identify associations and to assess the potential for
multicollinearity among the variables. As Table 8 shows, there is little cause to be
concerned about multicollinearity since the highest correlation (between LENGTH and

EDUC) is -.221.

Table 8 : Bivariate Correlation Coefficients of LENGTH, EDUC, and PROXIMITY

LENGTH EDUC PROXIMITY
LENGTH 1 .00
EDUC -.221 l .00
PRO)GMITY .051 .124 1.00

Preliminary analysis of my individual ENVIRO, TRANS, and SITE scores
produced high correlation values. The correlation values between ENVIRO, TRANS, and
SITE suggest a statistically significant relationship exists between the three variables. This
multicollinearity indicates that the combination of these variables may be used to test the
hypothesis presented in Chapter 1. Due to this high correlation, I proceeded with the
TOTAL value in my hypothesis testing.

.In addressing the hypothesis that I presented in Chapter I, I used regression and
t-test analysis to test the significance of the relationship between the risk perception levels
and socioeconomic factors. A t-test is used to compare the two study areas for the

hypothesized differences. Chi-Square analysis could also be used in studies of

46

Table 9: Bivariate Correlation Coefiicients of ENVIRO, TRANS, and SITE

ENVIRO TRANS SITE
ENVIRO 1 .00
TRANS .757 l .00
SITE .623 .823 1.00

comparison between two groups, but it is not the most usefirl test if the sample sizes are
small. Regression analysis was used to better understand what, if any, relationships
existed between the variables and risk perceptions. The first hypothesis concerns the total
risk perception scores of Mianrisburg and Dexter Village, while the last four hypotheses

concern the independent variables (SEX, EDUC, PROXIMITY, and LENGTH.)

Hypothesis 1: The residents of Miamisburg, Ohio will indicate a higher total degree of
risk perception concerning transportable hazardous materials than their Dexter Village,
Michigan, counterparts.

Due to Miamisburg's past, I surmised that the respondents of Miamisburg would
indicate higher levels of risk for all sets of issues. By comparing the two areas, I
attempted to difl‘erentiate between the levels of risk perceived. I turned to difference of
means t-test analysis for firrther explanation if a difference in total risk levels existed
between the two study areas. By placing the TOTAL variable into a t-test, I attempted to
distinguish between the two study areas and establish Miamisburg as having a higher level

of risk.

47

Table 10: Difference of Means T-Test between Miamisburg and Dexter Village
Respondents' TOTAL Scores

Mean Standard Deviation
Miamisburg 145.727 21.401
Dexter Village 135.706 26.372
Separate Variances DF= 26
Prob“: .302 = 1.054

The difl‘erence-of-means t-test has a calculated t (1.054) less than the critical t at
the alpha = .05 level; therefore, I can not accept the hypothesis that the Miamisburg
respondents perceive higher risk levels than their Dexter Village counterparts. Prior to
applying statistical tests, I suggested that the Miamisburg respondents had a heightened
sense of environmental awareness due to the 1986 derailment, but upon completing this

t-test I must indicate that this difference is not present.

Hypothesis 2: Female respondents will indicate higher levels of perceived risk than the

male respondents.

In chapter 11, I discussed the link between gender and perceived risk level. I
suggested that women would show higher perceived levels of risk due to their greater
sense of environmental responsibility. Again, a difference of means t-test was calculated
for the combined male and female respondents. This test was used to see if one gender

displayed a higher level of perceived risk over the other.

48

Table 11: Difference of Means T-Test between Female and Male Respondents TOTAL

Scores
Mean Standard Deviation
Male 114.00 42.922
Women 139.643 24.629
Separate Variances DF= 49
Prob.= .010 T: 2.674

This t-test has a calculated t value which is larger than the critical t at the alpha =
.05 level. Due to the larger calculated t value, I must conclude that female respondents
will indicate higher levels of risk than the male respondents. This analysis is consistent

with the literature introduced in Chapter II.

Hypothesis 3: Education is a significant indicator of perceived risk level, as those with
higher levels will perceive greater risk than those with less formal education.

As mentioned in Chapter H, past research has questioned whether those with
higher levels of education are more environmentally conscious than those less schooled. I
used a regression model relating TOTAL to EDUC to see if any relationship
exists.

By regressing TOTAL and EDUC, I have attempted to model a relationship
between an overall risk concern and the respondent's level of education. The regression
model explains only 3% of the variance and is not statistically significant at the set alpha=

.05 level.

49

Because there are other variables that may afi‘ect this relationship, I used a multiple
regression model with TOTAL, EDUC, and TOWN. I then proceeded with a model that
included TOTAL, EDUC, TOWN, and ET. The incorporation of the TOWN variable
allows for any relationship between the two study areas, education, and risk level to be
studied. VVrthin this intermediate model the EDUC variable is significant, but only 5% of

the variance is explained and the overall model is not significant.

Table 12: Regression Model of TOTAL and EDUC

Dependent Variable: TOTAL
Multiple R: .226

Multiple R2: .051

Adjusted Multiple R2: .030
Standard Error of Estimate: 35.847

Coefl' B T P
Constant 109.336 0.000 7.580 .000
EDUC 5.873 .226 1.571 .123

SS. Df M.S. F P
Regression 3169.991 1 3169.991 2.467 .123
Residual 59110.988 46 1285.021

I incorporated the ET variable within this model to ascertain if any discrepancies
between education, risk level, and town did exist. This regression explains 13% of
the variance within the model, while being significant at the alpha =.05 level. The EDUC

and TOWN coefficients are negative slants indicating that the education levels and town

50

Table 13: Regression of TOTAL, EDUC, and TOWN

Dependent Variable: TOTAL
Multiple R: .304

Multiple R2: .092

Adjusted Multiple R2: .052
Standard Error of Estimate: 35.440

Coefl‘ B Tol. T P
Constant 127.425 0.000 6.828 .000
EDUC 7.097 .273 .949 1 .871 .068
TOWN -15.092 -.209 .949 -l .426 .158

SS. Df M.S. F P
Regression 5760.566 2 2880.283 2.293 .1 13
Residual 56520.41 45 1256.009

designation are not positively related indicators of the total risk level. Due to the negative
direction of the coefiicients, I can not accept the hypothesis even though the model is
significant at the alpha = .05 level. Within this model are two equations, one for each
town. Both slopes for the towns are negative and Dexter Village's slope is greater
than Miamisburg. This negative slope indicates that the influence of education is more

pronounced in Dexter Village than in Miamisburg.

51

Table 14: Regression of TOTAL, EDUC, TOWN, and ET

Dependent Variable: TOTAL
Multiple R: .441

Multiple R2: .194

Adjusted Multiple R2: .139
Standard Error of Estimate: 33.774

Coefl‘ B Tol. T P
Constant 223.417 0.000 7.580 .000
EDUC -22.024 -.846 .075 -1.710 .094
TOWN -75.151 -1.042 .127 -2.743 .009
ET 17.749 .570 .041 2.356 .023

SS. Df M.S. F P
Regression 12090.133 3 4030.044 3.533 .022
Residual 50190.845 44 1140.701

Hypothesis 4: Proximity to the rail is significantly related to perceived risk, and those
living closer to the rail will perceive higher levels of risk than those at a further distance

fiom the railway.

After establishing the risk exposure area of two miles for both study areas, I can
incorporate the relationship between distance and risk levels. The proximity of the
respondents to the railway is expected to increase their perceived risk levels due to the
heightened concern for a deraihnent or dispersal.

The regression model of TOTAL and PROXIMITY explains less than one percent
of the variance and at the alpha=.05 level the overall model is not significant. From this

point, I will integrate the TOWN variable into the TOTAL and PROXIMITY regression

52

Table 15: Regression Model of TOTAL and PROXIMITY

Dependent Variable: TOTAL
Multiple R: .078

Multiple R2: .006

Adjusted Multiple R2: .000
Standard Error of Estimate: 36.370

Coeff B T P
Constant 122.813 0.00 1 1.293 .000
PROXIMITY 8.199 0.00 .548 .586

SS. Df M.S. F P
Regression 397.097 1 .097 .300 .586
Residual 64814.589 49 1322.747

model allowing for any distinctions between study area.

The regression model of TOTAL, PROXIMITY, and TOWN explained only 5%
of the variance, low tolerance levels signaling some multicollinearity, and is not significant
overall. Both of the independent variables are statistically significant. The opposite Of the
hypothesis is revealed in this model, as the TOTAL risk is increasing with the distance
from the railroad within the range Of study in both towns. The starting risk level is less in
TOWN=1, but the risk increases in each as the distance increases (See Figure 4) . The
idea that proximity is related to risk levels and breeds contentment is supported with these
results. I will next incorporate the PT variable to see if a relationship exists when these

variables are controlled.

53

Table 16: Regression Model of TOTAL, PROXIMITY, and TOWN

Dependent Variable: TOTAL
Multiple R: .302

Multiple R2: .091

Adjusted Multiple R2: .053
Standard Error of Estimate: 35.137

Coefl‘ B Tol. T P
Constant 147.425 0.00 9.417 .000
PROXIMITY 31.880 .303 .626 1.746 .087
TOWN -26.375 -.369 .626 -2. 121 .039

8.8. Df M.S. F P
Regression 5949.393 2 2974.697 2.409 . 101
Residual 59262.293 48 1234.631

This regression explains less than one percent of the variance within the model and
has only one variable within the model (TOWN). The tolerance levels are again low
indicating that multicollinearity exists within the model. The dummy variable (PT) allows
for a regression plane for each town, yet it is not statistically significant within this model.
The overall model is insignificant (p=.124.) The TOWN variable is significant yet the
slant of the slope is negative. The results presented in the previous model concerning the
increase in risk levels with distance is negated when incorporating the TOWN dummy
(PT). The increase in risk is at roughly similar rates in the two towns, but the starting
position of Town=1 is lower (see Figure 4). Although the significant relationship is not
related to the hypothesis it is worthy of note. Due to the irrsignificance of both models, I

can not accept the hypothesis that those living closer to the rail will perceive greater levels

54

Table 17: Regression Model Of TOTAL, PROXIMITY, TOWN, and PT

Dependent Variable: TOTAL

Multiple R:

.078

Multiple R2: .006
Adjusted Multiple R2: .000
Standard Error of Estimate: 36.370

Coeff B Tol. T P
Constant 186.987 0.00 4.781 .000
PROXIMITY -48.766 -.464 .037 -.647 .521
TOWN -51.120 -.715 .147 -l.994 .052
PT 46.191 1.028 .022 1.103 .276

SS. Df M.S. F P
Regression 7445 .416 3 2481.805 2.019 . 124
Residual 57766.27] 47 1229.070

of risk than those further away. The proximity of the rail lines to the residents in both

communities is very close and there may have been too little variation to detect any

relationships.

Figure 4: Comparison of TOTAL and PT Slope/Intercept Values

TOTAL

TOTAL = 147.425+3l.88*PROXIMITY-26.375*TOWN

Intercept =147.425 TOWN = l
/ TOWN = 0

Intercept = 121.05

 

 

PROXIMITY

55

Hypothesis 5: The length of residence is a significant indicator of risk levels as
respondents with greater lengths of residence will indicate higher levels of risk than those

with shorter length of residence.

The length of residence can be influential on the perceived level Of risk in that the
greater the length of residence, the greater the level of risk that will be perceived. This
should be evident in Miamisburg, as those who lived there ten years age may indicate

higher levels of risk due to their recollections and expoSure to the derailment.

Table 18: Regression of TOTAL and LENGTH

Dependent Variable: TOTAL
Multiple R: .320

Multiple R2: .102

Adjusted Multiple R2: .083
Standard Error Of Estimate: 34.340

Coefi B T P
Constant 150.311 0.00 14.550 .000
LENGTH -6.964 -0.320 -2.337 .024

SS. Df M.S. F P
Regression 6441.989 1 6441.989 5.463 .024
Residual 56604.011 48 1179.250

A regression model using TOTAL and LENGTH was executed to assess whether
the length of one's residence was influential in the level of risk perceived by the
respondent. The overall model was significant, yet the coefficient for the LENGTH

variable is a negative slant. This negative slant may suggest that the greater one's length

56

of residence, the lower one's perceived risk levels due to the continued transportation of
such materials without any type of mishap. This may be descriptive of Dexter Village's
attitudes as the residents did not harbor any personal experiences with hazardous
materials. Again, this does not support the hypothesis. At this point, I can not accept the
hypothesis that the greater the length of time a respondent resides in the area, the higher
the perceived level of risk.

I have continued to model a regression to see if any differences exist between the
two study areas. For the next model, I incorporated the TOTAL, LENGTH, and TOWN
variables into a regression. By adding the dummy variable (TOWN) into the model, I can
assess whether any differences exist between the Dexter Village and Mianrisburg

respondents regarding risk level and length of residence.

Table 19: Regression of TOTAL, LENGTH, and TOWN

Dependent Variable: TOTAL
Multiple R: .410

Multiple R2: .168

Adjusted Multiple R2: .114
Standard Error of Estimate: 33.760

Coefi‘ B Tol. T P
Constant 170.500 0.00 9.373 0.00
LENGTH -7 .208 -.331 .996 -2.435 .119
TOWN -12.962 -. 183 .996 -1.343 .186

88. Df M.S. F P
Regression 8534.140 2 4267.070 3 .679 .033

Residual 54511.860 47 1 159.827

57

Figure 5: Comparison of TOTAL and LENGTH Slope/Intercept Values

TOTAL = 170.500 - 7.208 * LENGTH - 12.962 * TOWN

TOTAL Intercept = 170.500

VN=0

Intercept = 157.538 TOWN = 1

 

 

LENGTH

The tolerance levels within this model are very high and the overall model is
statistically significant; yet, the independent variables Of LENGTH and TOWN were not
significant. Less than 12% ofthe variance was explained in this model. Again, I will
incorporate another dummy variable into the model (LT) which describes the length of
residence multiplied by the town dummy variable. If the LT were to be significant, the
variable would describe the two study areas on different regression planes (See Figure
5).

This regression explains less than 12% Of the variance within the model but is
statistically significant at the alpha = .05 level. None of the independent variables within
the model are significant. The two towns have the same slope yet TOWN=O has a higher
intercept suggesting that as length increases the TOTAL risk level decreases. Although

this finding is not in the direction of the hypothesis, it is worthy of note.

58

Table 20: Regression of TOTAL, LENGTH, TOWN, and LT

Dependent Variable: TOTAL
Multiple R: .410

Multiple R2: .168

Adjusted Multiple R2: .114
Standard Error of Estimate: 33.760

Coefl‘ B Tol. T
Constant 126.367 0.00 3.389
LENGTH 6.933 .318 .073 .638
TOWN 13.206 .186 .196 .612
LT -8.456 -.749 .059 -l.352

S.S. Df M.S. F
Regression 10618.701 3 3539.567 3.106
Residual 52427 .299 46 1 139.724

Table 21: Review of Hypothesis Outcomes

1. Miamisburg Risk Levels > Dexter Village Risk Level Reject
2. Females will indicate higher levels of risk than Males Accept

3. Higher Education Levels = Higher Risk Levels Reject
(the opposite prevailed)
4. Closer Proximity to the rail = Higher Risk Levels Reject

5. Greater Length of Residence = Higher Risk Levels Reject

.000
.526
.544
.183

.036

Chapter V:

Discussion

The Objectives Of this study were to measure the perceived risks in two
communities, assess any relationships, and to discuss if demographic factors were
influential in the perception process. The study's hypotheses were used to propose that
socioeconomic factors were influential in how risk is perceived and that one of the
communities would indicate higher levels of risk due to their past experience with a train
derailment.

Past risk perception research has focused on the controversies surrounding facility
siting and disposal. Awareness of industry's past practices has created increased
regulatory oversight and community concern over potential environmental impacts.
Anxiety over disposal practices has focused on what long term threats are posed by
hazardous materials not only to the natural environment but to human health.

The purpose of this study was to analyze and assess risks perceived by two
Midwestern towns in association with the transportation of hazardous materials and to
determine if these groups differed in their perceptions. Dexter Village, Michigan, and
Miamisburg, Ohio, were chosen for a comparative assessment of two areas with similar

socioeconomic compositions and are currently encountering hazardous materials traveling

through their limits.

59

60

The study results established a ranking of risk levels which were used to test the
hypothesis presented in the first chapter. The results did not concur with all of the stated
hypotheses. The perceived levels of risk were much more similar than dissimilar between
the two study areas. However, the results did indicate a consensus among respondents
that hazardous materials transportation mishaps are something that those in planning
positions should work to mitigate.

Literature cited in this study established the concern for hazardous material
processes including treatment, storage, and disposal. Whereas Slovic has looked at both
siting issues in the Yucca Mountains and highway transportation in Oregon, this study
identified the risks perceived with respect to transporting hazardous materials by rail. The
use of risk factors either jointly or individually, such as the highest risk to community
water supplies (Great Miami and Huron Rivers) and rail conditions, can be targeted by
planners as those areas in which to better inform the public.

Risk perception as an additional facet in the planning process is vital for the
continuation of effective policy. The necessity for local communities to capably handle the
issues facing their residents and to have intact contingency plans in the event of a mishap
is becoming more evident. Miamisburg experienced this type of incident first hand and
witnessed the ramifications of the lack of communication between Local, State, and

Federal officials.

61
Qpnplpsions and Recommengtipps

The commencement of this project brought several ideas and generalizations about
how the respondents of these two communities may react and perceive risks associated
with the transporting of hazardous materials. I knew the history of the Miamisburg area,
of the ramifications of the derailment, and the attitudes toward the DOE Mound facility. I
knew very little about Dexter Village beyond the fact that it was a quaint town tucked
outside of Ann Arbor. Both towns had one commonality though, that rail trafic is going
through their communities carrying hazardous materials.

The conclusions drawn fiom this study were surprising. The degree to which the
two communities agreed on the highest and lowest risks to their communities was
hypothesized to be significantly different. I expected to see more dissension than
agreement. I suspected that Miamisburg respondents would indicate extremely high levels
of risk throughout the survey while Dexter Village might have more variations. Instead,
the study presented a high degree of agreement between Dexter Village and Miamisburg.
Following is a more detailed discussion of the study’s conclusions. It must be
acknowledged, however, that there may have been serious problems with the clarity of the

questionnaire that limits these conclusions.

Caperal Environmental Isspes
The overall ranking of the general environmental issues by both towns were
shown to be relatively similar. The ranking of the highest risk to human health "Toxic Air

Pollutants" and "Industrial Discharges" as the highest risk to the environment were

62

established concerns demonstrated by the respondents. I suggest that the air pollutants
and discharges were perceived as the highest levels in Miamisburg due to the phosphorous
cloud from the derailment and the current remediation ongoing at the DOE Mound
facility. Dexter Village respondents may have indicated air pollutants and discharges due
to the unknown nature of the materials. Air pollution is an unknown risk to the average
human due to the relative invisibility Of the materials. Industrial discharges conjure up
images of Love Canal and the threat of toxins polluting our water supplies and homes.
This study analyzed two Midwestern towns on the basis of their perception of risk
in transporting hazardous materials. The similarities that exist could be fostered by the
media and the increasing attention given toward environmental hazards. The highest risk
levels identified by the respondents could be addressed by community planners and

implemented into community educational planning.

Transportation Risk Factors

Shierbl identified transportation risk factors in his study of inter-agency perception
regarding the transportation of hazardous materials (1994). By using the transportation
risk factors in this study, I could assess the perception of risks associated with the
transportation process. The ranking of risk factors was more diverse in this section. The
respondents identified "Condition of the Rail and Tank Cars" as the highest risk factor,
while "The Time of Day when Rail Shipments are Transported" was chosen the least risky.

The recognition Of rail conditions as the most risky transportation factor is

understandable in Miamisburg, as the deraihnent was caused by a kink in the tracks. The

63

Dexter Village respondents concern for the rail condition is natural. Ultimately that is the
most important factor in the transporting of any material over any distance. The need for
safe, reliable rail routes is a must for the continued transportation of hazardous materials.
The "Time of Day" as being the least risk is not surprising, as very few people
ever interact with rail transportation. Most rail routes have very few public crossing
points; therefore, the old cliché' "Out of Sight, Out of Mind" is the sentiment of most
respondents. An interesting aside though is that a few Dexter respondents expressed
concern over the trains going through the Village at night. For highway transport, many
communities have time periods in which materials may be transported (e. g. not during
school hours), but I do not know of any current or firture plans for time designations for

transporting by rail in either Miamisburg or Dexter Village.

Site Risk Factors

By choosing two study areas with comparable topography, I could assess
perceptions dealing with site specific factors which were in both towns. The respondents
of both Dexter Village and Mianrisburg indicated that their own "homes" were at the
greatest risk. This ranking is due to the proximity of the towns' population densities to the
rail (side by side.) I suggest that this concern was also conjured on the fear of the
unknown in Dexter Village, as the threat of something infiltrating our private homes is
relatively scary. The high number of Miamisburg respondents who indicated that they
were in Miamisburg during the derailment had experience with this type of risk and
therefore indicated high risk levels.

64

The respondents' indicated that the "town population" was at the lowest risk from
any type of incident. I suggest that this is due to the evacuation of the entire town in
Miamisburg and the surrounding communities after the derailment. The thinking behind
this lower level risk may be that the "town population" is all encompassing and most spills
or accidents would likely be smaller scale.

The relevance of the site risk factors can be attributed to the awareness of the
vulnerability of the communities. Miamisburg is consistently given as an example, but the
1986 derailment demonstrated how a mishap can disrupt everyday life and possibly cause
long term effects. This type of analysis and recognition of local issues is instrumental in
developing contingency plans within the community.

Hypotheses

The hypotheses presented in Chapter I were statistically tested in Chapter IV. Of
the five hypotheses presented, I could only accept one. My hypotheses were based on the
socioeconomic/perception findings of past literature. The literature basis of environmental
risk perception is still in the infant stages, but I submit this study's results as a basis for
further research.

Prior to executing this research, I had definite ideas about the risk awareness in
each of the study areas. I anticipated Miamisburg having a heightened level and Dexter
Village exhibiting some awareness. The testing of my hypotheses dispelled this
preliminary thinking. The education and length of residence relationships that I
hypothesized were actually the opposite of the analyzed results. The length variable

indicated a negative slope within the regression model suggesting that one may negate any

65

types of risk due to the continued transportation of materials through the communities
with little or no problem. The proximity variable within my regression models suggested
that the risk levels began to increase as the distance began to increase indicating that those
further from the transportation corridors may view transportation as more risky. The
broad range of environmental activities in communities today, such as recycling, has
created a greater awareness than may have been anticipated. The demographic

relationships that I suggested were also not statistically significant.

Willem

Upon completion of this study, I began to reflect on the overall structure of the
study and address issues that could be clarified for any firture attempts at perception
studies. The need for perception studies remains vital for the formulation of public policy;

therefore, the methods that reveal public attitudes clearly and accurately are crucial.

The urve

The success of a perception study is dependent on the responses received. If I
were to attempt another study analogous to this one, I would implement a survey
distribution that would involve more person-to-person interaction. The theory is that the
more personal interaction, the greater the number of completed responses (Dillman 1978).
This interaction can be achieved by telephone interviews or focus groups.

With the advent of telemarketing, the telephone is viewed as one route to survey a

representative sample about the issue of choice. The advantages of this type of survey

66

include immediate responses, a clarification of any questions, and the ability to better
inform the respondent of any questions or concerns regarding the survey. The
disadvantages of a phone survey are the expense (long-distance), time intensive nature,
and the potential need for additional staff, not to mention the disruption of activity for the
respondents. The telephone method would appear to be quite harmless, but I suggest that
the intrusive manner of phone surveying may not result in a greater number of completed
responses. This method is usefiil but may not be the best approach for perception studies.

The Focus Group method is one technique that is commonly used by governmental
entities for environmental policy creation. The groupa generally include a mixture of
citizen action groups, municipalities, technical people, industry, university Officials, and
state agencies. One example of the use of such group in an environmental study is the
effort by the Ohio Low-Level Radioactive Waste Facility Development Authority. The
Authority currently uses a Focus Group in their attempt to successfully site a Low-Level
Radioactive Waste Facility within Ohio. Within a perception studies context, a focus
group could utilize a representative sample of both study areas and survey their
perceptions of risks from transportable hazardous waste.

The advantages of the Focus Group include clarification of questions, immediate
responses, and a better understanding of the project and what is to be gained by such an
undertaking. The disadvantages are the difficulty in assembling a representative sample,
scheduling conflicts, and potential lack of interest. For any firture perception analysis, I

suggest that the focus group method may yield the most results.

67

One additional survey method that could be used for perception studies is that of
the web page. The dramatic increase in Internet usage has established another route by
which a survey can be distributed. By creating a web site, virtually anyone with a
computer and internet service can log on and complete a survey. This method is very

useful when demographics are not a component of the study and it is also inexpensive.

Qupsp'pnnaire Struflre

The creation of any future perception studies would result in various changes in
my survey instrument. The questionnaire that I distributed in Dexter Village and
Miamisburg included Likert style scaling with closed-ended questions. This type of format
is useful in gauging perceptions, but one person may indicate a "High" response as a "4"
while another respondent may consider a "High" perception level a "6. " The respondents
would then perceive the risk at the same level but indicate two difi'erent numbers of risk.
The more appropriate scale may be: NO Risk, Low Risk, Medium Risk, High Risk,
Extreme Risk, perhaps comparing to standards (being struck by lightening, being in an
auto accident, etc.) This scale would allow the respondent to identify the wording with
their exact perception level. At the very least, some questions should be included that
allow comparison of levels. For example, there is no reason to expect a difi‘erence between
Miamisburg and Dexter Village on perception of risk of the United States facing some
particular environmental problems. Other values could then be scaled relative to the

individual's average values on those items.

68

I suggest, too, that various wording changes be made before re-administering this
type of perception question. I submit that I may have created directions and
environmental risk factors that may have been difficult to understand for the average lay
person. The directions could be stated, "Given your understanding of how potentially
hazardous materials are currently handled and regulated, what level of risk do you believe
to be associated with: Air pollutants from industry, Industrial Discharges to Surface
Water, etc? “With this approach, I may have had more residents respond to the
questionnaire rather than discarding them. A questionnaire that is clearer and involves
more risk factors within the study area will probably yield higher response rates and an
overall interest within the study areas. Table 22 is a shortened, revised sample. The "NO
Risk to Extreme Risk" scale can be converted to a numerical scale (1 through 5) for input
into statistical analysis. The same conversion can take place for the percentage likelihood
(between .00 and 1.00).

The demographic section of the questionnaire is included to ascertain what
socioeconomic variables may correlate to certain perceived risk levels. I included in this
survey's questionnaire common questions concerning Racial Identification, Household
Composition, and Yearly Household Income. The inclusion of such categories may have
been a deterrent for some potential respondents because this type Of information is very
personal. I did not need to include the Racial Identification since I knew that both study
areas are almost completely homogeneous. I did include the Income and Household

composition categories though to assess what relationships may exist between risk levels

69

and these variable. Due to the lack of responses containing this information, I decided to
withdraw this type of analysis fi'om the research project.

An additional facet that I would address for firture perception studies is the
distance from which to survey an exposure area. The wide-spread use of the information
super highway and twenty-four hour news networks creates a broader exposure for the
dissemination of news. For example, the 1986 Miamisburg derailment was broadcast in
Southern Ohio and probably few other places. The 1996 Wisconsin derailment was
broadcast on the National networks. Due to this increased exposure, I would try to assess
perceptions at greater distances from the study areas. For the Dexter Village scenario, I
would distribute surveys in Ann Arbor, Brighton, Metropolitan Detroit, and Jackson.
The Miamisburg exposure area would include Dayton, Springfield, Columbus, and

Cincinnati. I would also attempt some random surveys throughout the United States.

70

Table 22: Sample of Reconstructed Survey

Part 1: Environmental Issues Concerning Hazardous Waste Transportation

Given your understanding of how potentially hazardous materials are currently
handled and regulated, what level of risk do you believe to be associated with:

Risk flsk to Human Health
1. Air Pollutants from Industry NO Risk Low Medium High Extreme
2. Industrial Discharges to Surface Water No Risk Low Medium High Extreme
3. Contaminants from Hazardous Waste NO Risk Low Medium High Extreme

Storage Sites

4 Drinking Water fi'om Tap NO Risk Low Medium High Extreme
M Risk to the Envirpnment

5. Air Pollutants from Industry NO Risk Low Medium High Extreme

6. Industrial Discharges to Surface Water NO Risk Low Medium High Extreme

7. Contaminants from Hazardous Waste
Storage Sites
Drinking Water from Tap

No Risk Low Medium High Extreme

.°°

No Risk Low Medium High Extreme

Part II: Hazardous Materials Transportation

The following is a list Of risk factors attributable to hazardous material
transportation. What level of risk do you believe to be associated with each factor:

M Risk Level
1. Volume of Materials Transported No Risk Low Medium High Extreme
2. Distance Materials Are Transported No Risk Low Medium High Extreme
3. Condition of the Rail and Tank Cars NO Risk Low Medium High Extreme
4. Weather Conditions During Transport No Risk Low Medium High Extreme
5. Population Density Along Railroad No Risk Low Medium High Extreme
6. Natural Resources Along Railroad NO Risk Low Medium High Extreme
7. Emergency Response Along Railroad NO Risk Low Medium High Extreme
8. Type of Chemicals Transported No Risk Low Medium High Extreme
9. Time of Day During Transport No Risk Low Medium High Extreme
10. Chemical Storage During Transport No Risk Low Medium High Extreme

71

Part III: Hazardous Materials Within Dexter Village/Miamisburg

Given your understanding of the amount, frequency, and type of hazardous
materials moving through your community by rail, what do you think is the likelihood of
an accident harmfirl to each of the following in the next ten years. Give any value fiom
0% to 100%.

R_is__k Likelihood

1. The Huron/Great Miami River
2. Your Home
3. The Dexter Village/Miamisburg Population

What do you think is the likelihood of the following:
4. A Chemical Dispersal In Dexter Village/Miamisburg
5. Evacuation fi'om your Home

Part IV: Background Information

Age Sex: M F
Length of Residence: Years
Educational Level Attained: Months

Less Than High School
High School or Equivalent
Some College

College Completed
Technical School
Graduate School

Are you or any member of your household currently active in any environmental
organizations? Y N

(Dexter Village only)
Are you aware of Conrail transporting hazardous materials through Dexter Village? Y N

(Miamisburg only)
Did you live in Miamisburg during the 1986 derailment? Y N
Were you Evacuated? Y N

Do you feel the elected and regulatory Officials are actively pursuing ways to mitigate any
type of transportation incident which might involve hazardous materials? Y N

72

My reflection of this study has suggested some areas I would change if this study
were repeated. I suggest that perception studies, such as this, are vital to the ever
evolving creation Of environmental public policy. As the volume of perception study
continues to increase, more literature will prevail suggesting new socioeconomic variables
that may influence perceptions. I submit this study to the growing volume of literature
and look forward to the attention that hazardous materials transportation risks will

continue to receive.

APPENDIX A

73

 

MlCHlCAN STATE
U N I V E R S | T Y
Dear Respondent:

The Department of Geography at Michigan State University is conducting an exploratory study of
risks associated with the transportation of hazardous wastes on Ohio and Michigan railroads.

This questionnaire is being administered to a cross section Of Dexter Village, Michigan, and
Miarrrisburg, Ohio, inhabitants in order to learn about and suggest improvements in the way
hazardous waste transportation programs and policies are implemented in both states. It will take
approximately 15 nrinutes to complete this questionnaire.

A report describing the study's results and potential implications for public policy will be
generated from the Obtained information and available for anyone to review. Responses to the
questionnaire will be analyzed on a group basis only and will remain totally confidential. Please
return the completed questionnaire in the envelope included in this package by Friday, September
1, 1995. Thank you for your time and assistance in this study. Your participation is essential in
obtaining useful results.

Sincerely,

WW5
Beth A Myers

Graduate Student

W494i @3032}?

Dr. Bruce Pigozzi
Department of Geography
Project Advisor

APPENDIX B

74

MICHIGAN STATE

UNIVERSITY

September 9, 1995
Dear Respondent:

One month ago a questionnaire asking for your views on the transportation of hazardous materials
by rail through Dexter Village/Miamisburg was distributed to you. If you have completed the
questionnaire already, please accept our sincere thanks. If not, could you please return it today?
The questionnaire was sent to only a small representative sample; therefore, it is important that
your views are included in this study if we are to represent the people Of Dexter Village’s/
Miamisburg's views accurately. .

If you did not receive a questionnaire or have mislaid it, I have enclosed another copy and
stamped envelope within this packet.

Sincerely,

WWVé’
Beth A. Myers

Graduate Student

 

Dr. Bruce Pigozzi
Department of Geography
Project Advisor

APPENDIX C

75

Part 1. Environmental Issues Concerning Hazardous Waste Transportation

A. Below are issues pertinent to the environmental concern and the subsequent
transportation of hazardous materials. Please circle the number that best describes you
personal understanding of each issue and if it is deemed a risk either to the environment or
to human health. The response scale ranges from O= NO Risk to 6: High Risk.

Risk to Risk to
Risk Factor Human Health the Environnrent
None--Medium--High None--Medium--High
1.Airpollutantsincluding 0123456 0123456

sulfur dioxide, total suspended
particulates, carbon monoxide, nitrogen,
oxides, ozones, and lead.

2. Toxic air pollutants including dioxin 0 1 2 3 4 5 6 O 1 2 3 4 5 6
3. Industrial dischargestosurface water 0 l 2 3 4 5 6 O l 2 3 4 5 6
4. Contaminantsrcleasedfrom 0 l 2 3 4 5 6 O l 2 3 4 5 6
hazardous waste sites that are
actively being managed
5. Accidental releases of chemicals 0 l 2 3 4 5 6 0 l 2 3 4 5 6
6. Radiationfromsourcesotherthan O l 2 3 4 5 6 0 l 2 3 4 5 6
indoor radon
7. Drinking water fi'om tap O 1 2 3 4 5 6 O 1 2 3 4 5 6

8. Workerexposuretochemicals 0 1 2 3 4 5 6 0 l 2 3 4 5 6

76
B. Part H. Hazardous Materials Transportation

The following is a list of risk factors attributable to hazardous material transportation. Please circle
the number that best describes your personal feeling of the overall "risk" represented by each issue.
The response scale ranges from 0=No Risk to 6=High Risk. ie. The following question should be
read as, "The type of materials being transported represent which level of risk?"

Risk NO Risk High Risk
1. Thetypeofmaterialsbeingtransported . 0 l 2 3 4 5 6
2. Volumeofmaterialsbeingtransported 0 l 2 3 4 5 6
3. Emergency response capability along transport route. 0 l 2 3 4 5 6
4. Distancethehazardous materialsarebeingtransported O 1 2 3 4 5 6
5. Population density along the railroad. O l 2 3 4 5 6
6. Natural resource density andtypealong the railroad. 0 l 2 3 4 5 6
7. Conditionoftherailandtankcars. O l 2 3 4 5 6
8. The time of day when rail shipments are transported. 0 l 2 3 4 5 6
9. Weather conditions during transport. 0 1 2 3 4 5 6
10.Typeofstoragewithintankcarduringtransport. 0 1 2 3 4 5 6

III. Hazardous Material Transportation within Dexter Village/Miamisburg

The following is a list of risk factors attributable to hazardous material transportation by
railroad within Dexter Village/Miamisburg. Please circle the number that best describes

your personal feeling of the overall "risk" represented by each issue. The response scale
ranges fiom 0=No Risk to 6=High Risk.

Risk No Risk High Risk
1. TheHuron/GreatMiamiRiver 0 1 2 3 4 5 6
2. The respondent's home 0 1 2 3 4 5 6
3. The Dexter Village/Miamisburg Population 0 1 2 3 4 5 6
4. Chemical dispersal in Dexter Village/Miamisburg O l 2 3 4 5 6
5. Tank car derailment or accident 0 1 2 3 4 S 6

77

Part IV. Background

Age
Children
Sex: M F Length Of Residence: Years
Educational Level Attained: Months
Some High School
High School or Equivalent
Some College
College Completed
Technical School
Graduate School
Racial Identification: Yearly Household Income
Afiican-American $0-7,000
Asian $7, 000-1 5,000
Caucasian $15,000-22,000
Latino $22,000-30,000

Native American $30,000-40,000
“0,000-50,000
$50,000-60,000

$60,000 and greater

Are you or any member of your household currently active in any environmental
organizations? Y N

Are you aware Of Conrail transporting hazardous materials through Dexter Village? Y N

‘ Did you live in Miamisburg during the 1986 derailment? Y N
Were you evacuated? Y N

Do you feel the elected and regulatory Officials are actively pursuing ways to mitigate any
type of transportation incident which might involve hazardous materials? Y N

APPENDIX D

78
m r n in ' ' r hi
I was working in Miamisburg on July 8, 1986 (train derailment) and had to leave.

My family has had lung problems and bad health ever since the derailment, and I believe if
officials would have told the public more about the problems with the type of chemicals,
but instead all they told us was it was not a health problem.

I am in the area that was asked to leave at the time of the derailment, but far enough away
to be affected with the smoke drifting my way.

Precautionary measures were only temporary after the 1986 derailment, train speed is back
up and cherrrical cars are not dispersed throughout the total length of the train. The DOE
has NOT been open and honest with Miamisburg residents about CHEMICAL and
RADIATION RELEASES from the "former" nuclear weapons research facility.

I feel that too much hazardous waste is being transported secretly and it all boils down to
money. Companies will ship their waste long distances to avoid the high cost of disposal
within their locality.

I am more concerned about Mound and what it has done to the water, etc. Also, space
crafts going in space are surely not good on ozone either.

My belief is that as long as these companies are making money, they don‘t care about the
environment. It is really too bad because if they don't start to care about the Earth it may
never heal.

Why did you ask these questions?

‘ Comments frpm resppndgnts in DQflgt‘ Village, Michigan

I was not aware that hazardous materials are being transported so close to my house. I
don't like it. It all feels like high risk to me. It's scary.

I am concerned that they are producing hazardous waste to begin with- much less moving
them around the country.

Thanks for the opportunity to fill this out.
Until this survey, 1 had not been aware of anything to do with this subject.

Most of the freight trains come through at night, as far as I can determine. Those 1 do see
during the daylight hours don't seem to have warning signs on the cars.

79

I will not vote for anything that will raise my taxes - most of you- waste our money with
high salaries and big committees to do nothing.

Local government has taken no interest in the Conrail issue, as the train runs thru Chelsea,
Dexter, Ann Arbor, and Ypsilanti. However, the local fire departments have a haz mat
team, they also work with Wayne Co.

Great Lakes region is especially vulnerable and requires strict protection. Small towns
cannot support the infrastructure necessary to deal with a major event.

LIST OF REFERENCES

LIST OF REFERENCES

Arbuthnot, Jack. 1977. " The roles of Attitudinal and Personality Variables in the
Prediction of Environmental Behavior and Knowledge". Environment and
Behavior 9, 2: 217-231.

Arcury, TA. and Scollay, SJ. and TR Johnson. 1987. "Sex Differences in Environmental
Concern and Knowledge: The Case of Acid Rain". Sex Roles 16: 463-472.

Backstorrrr, Charles H. and Gerald Hursch. 1963. Survgy Rgssargph. Chicago, Illinois:
Northwestern University Press.

Borden, Richard J., and Janice L. Francis. 1978. "Who Cares about Ecology: Personality
and Sex Difi‘erences in Environmental Concern". J oumgl of Personality 46,
1:3 5-39

Borden, Richard J. and Andrew P. Schettino. 1979. "Deterrrrinants of Environmentally
Responsible Behavior". Journgl Of Environmental Education 10, 4:35-39.

Borsky, RN. 1961. "Community Relations to Air Force Noise-Data on Community
Studies and their Interpretation." Dayton, Ohio: Wright-Patterson Air Force Base.

Bruvold, William H. 1973. "Belief and Behavior as Determinants of Environmental
Attitudes" Environment and Behavior 5, 2: 202-18.

Bunting, Trudi E. and Leonard Guelke. 1979. "Behavioral and Perception Geography: A

Critical Appraisal." Annals of the American Association of Geographers
69:448-462.

Burton, Ian and Robert W. Kates. 1964. "The Perception of Natural Hazards in Resource
Management." Natural Resources Joum_a_l. Albuquerque, New Mexico:
University of New Mexico School of Law.

Buttel, Frederick H. and Donald E. Johnson. 1977. "Dimensions of Environmental

Concern: Factor Structures, Correlates, and Implications for Research". Journal
of Environmentafl Education 9, 2:49-64.

Calef, Wesley. 1950. "The Winter of 1948-1949 in the Great Plains". Annals of the
Associapion of Ameripan Geographers. 40: 267.

 

80

81

Clapharn, W.B., Jr. 1990. "Some Approaches to Assessing the Environmental Risk in
Siting Hazardous Waste Facilities". The Environmental Professional 12, 1:32-39.

Clarke, Lee. 1989. "Acceptable Risk? Making Decisions in a Toxic Environment".
Berkeley: University of California Press.

Commoner, Barry. 1988. "Acceptable Risks: Who Decides?" Hagper's Magazine. May
1988.

Constantini, Edmond and Kenneth Hanf. 1972. "Environmental Concern and Lake
Tahoe". Environment and Behavior June.

Craik, Kenneth H. and Ervin H. Zube. 1976. Perceivingéinvironmeml Ouafly. New
York: Plenum Press.

Crowe, MI. 1968. "Towards a Defirritional Model of Public Perceptions of Air
Pollution." Journal of Air Pollution Control Association. 18: 154-157.

Cunningham, William P. 1994. Unoorstanding Our Environment. Dubuque, Iowa: Wm.
C. Brown Publishing.

Dear, Michael. (1992). "Understanding and Overcoming the NIMBY Syndrome." Journal
of the Americap Planning Association. 58: 288-292.

Dillman, Don A 1978. Mail and Telephone Surveys, The Total Design Method. New
York: John Wiley and Sons.

Douglas, Mary and Aaron Wildavsky. 1982. Risk and Culture. Berkeley: University of
California Press.

Dunlap, Riley E. and Kent D. Van Liere. 1978. The "New Environmental Paradigm."
Iournal of Environmentol Education 9, 4:10-19.

Flora. (1956)1_1gilstorms of the United Stotos_.

Goodey, Brian. 1974. Imogos of Plgoo; Essfls on Environmontal Peroeption,

Communication and Education. Birmingham, England: The Centre for Urban and
Regional Studies.

Gregory, Robin and Robert Mendelsohn. 1993. "Perceived Risk, Dread, and Benefits".
Risk Analysis 13, 3: 12-19.

82

Kunreuther, Howard and Ruth Patrick. 1991. "Managing the Risks of Hazardous Waste".
Environment 13, 4: 21-26.

Lowenthal, David. 1967. "Environmental Perception and Behavior" Department Of
Geography research paper 109. Chicago, Illinois: University of Chicago.

Marsh, Catherine. 1982. The Sowey Method: The Contribution of Surveys to
Sociologioaj Explanation. Boston: George Allen and Unwin.

McNeil, Sue and Sci-Chang Oh. 1991. "A Note on the Influence Of Rail Defects on the

Risk Associated with Shipping Hazardous Materials by Rail". Risk Analysis 11, 2:
14-20.

Morgenstem, Richard and Stuart Sessions. 1988. "Weighing Environmental Risks: EPA's
Unfinished Business". Environment 30, 6: 14-39.

National Transportation Safety Board. 1981. Status Of Department Of Transportation's
Hazardous Materials Regolatog Program. Department of Transportation.

Noe, Francis P. and Rob Snow. "The New Environmental Paradigm and Further Scale
Analysis". Journfl of Environmental Education 26: 20-26.

Roberts, L. 1990. "Counting on Science at EPA". Science 2492616.

 

Saccomanno, FF. and 1H. Shortreed. 1993. "Hazmat Transport Risks: Societal and
Individual Perspectives". loom Of Transpogtaiion Engineering 119, 2: 177-188.

Schahn, Joachim and Erwin Holzer. 1990. " Studies of Individual Environmental Concern".
Environment and Behgvior 22, 6: 767-786.

Sheskin, Ira M. 1985. " Survey of Research for Geographers". Washington, DC:
American Association of Geographers.

Slovic, Paul. 1986. "Informing and Educating the Public about Risk". Risk Analysis 6, 4.

 

Slovic, Paul. 1987. "Perception of Risk". Science 236: 280-85.

 

Slovic, Paul, Mark Layman, Nancy Kraus, James Flynn, James Chalmers, and Gail Gessell.
1991. "Perceived Risk, Stigma, and Potential Economic Impacts Of a High Level
Nuclear Waste Repository in Nevada". Risk Analysis 11, 4: 683-696.

83

United States Bureau of the Census. 1990. 1990 Michigan State Census.
United States Bureau of the Census. 1990. 1990 Ohio State Census.

Van Liere, Kent D. and Riley E. Dunlap. 1981. "Environmental Concern". Environment
and Behavior 13, 6: 651-675.

 

Wall, G. 1972. "Public Response to Air Pollution in Sheffield". International Geographical
Union, Commission on Man and Environment. Calgary Symposium. July 23-31.

White, G. 1964. Papers on Flood Problems. Chicago, Illinois: University of Chicago
Department of Geography.

“v

Om

"IIIIIIIIIIII