meats

2M4
6¢<&SZ?>

Date

0-7 639

IJIHRAJCY
Michigan State
University

This is to certify that the

thesis entitled
DETERMINING THE S-DABILITY
OF VARIOUS FIBER TYPES
presented by

Sarah Elizabeth Walbridge

has been accepted towards fulfillment
of the requirements for

M. 8. degree in mm Justice

on )M

‘/,Major pldfessor

Max/03

MS U is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p.15

 

 

DETERMINING THE SHEDDABILITY OF VARIOUS FIBER TYPES
By

Sarah Elizabeth Walbridge

A THESIS
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
School of Criminal Justice

2003

ABSTRACT
DETERMINING THE SHEDDABILITY OF VARIOUS FIBER TYPES
By
Sarah Elizabeth Walbridge

The shedding quality of a fabric has utility in forensic science. Fibers are
frequently recovered as evidence in criminal cases. The application of the shedding
potential of a garment is thus significant. A standard method for determining the
sheddability index of various fiber types was developed. The low-adhesive backing of a
3” x 3” 3M© post-it note was utilized to simulate a ‘natural’ shedding of the fabric. A
raw mean shedding index was assigned for the thirty—one fabrics tested by this method.
Based on this data, fabrics were categorized from high to low shedders. Fiber-transfer
studies with nineteen of the tested fabrics as donors and white/black felt as recipients was
performed and compared to the previously determined shedding index categories. Two
simulated contacts, a handshake (no contact), and a hug (friendly contact) were used as
transfer conditions. The hug transfer was repeated three times to calculate a mean.

Overall, the method developed for determining the sheddability of a fabric is
efficient, reproducible, and reliable. The post—it note method results correlated well with
the shedding potential. The rank correlation coefficient (R:0.753) indicates that Raw
Mean Shedding Index and Mean Transfer are positively correlated. This supports the
hypothesis that determining the sheddability of a fabric is a useful measurement for

evaluating fiber transfer.

ACKNOWLEDGMENTS

The author wishes to thank all those who have encouraged and supported the
success of this research project. The author owes a large thank you to Dr. Jay Siege] for
the encouragement, patience, and guidance through this project and many occasions
beyond. The author is equally grateful to Max Houck for his assistance, feedback, and
ideas which were constantly requested and always thoughtfully provided. The author is
grateful to Rick Myers for participating in the transfer study. The author also extends a
thank you to her family who has waited patiently for the completion of this research
project with continuing support. Lastly, thank you Chris for the many Sundays of

unassisted laundry duties.

iii

TABLE OF CONTENTS

List of Tables ......................................................................................... vi
List of Figures ....................................................................................... vii
Chapter 1
Introduction ............................................................................................ 1
Chapter 2
Literature Review ..................................................................................... 7
Transfer and Persistence .......................................................................... 7
Secondary Transfer ............................................................................... 10
Fabric Sheddability ............................................................................... l 1
Hypothesis of Current Study .................................................................... 13
Chapter 3
Methods and Materials ............................................................................... 15
Fabrics Chosen for this Research ................................................................. 15
Materials Utilized for Determining Shedding Potential ....................................... 16
Method Utilized for Determining Shedding Potential ......................................... 17
Materials Utilized for Transfer Studies .......................................................... 19
Method Utilized for Transfer Studies ............................................................ 19
Handshake Transfer ................................................................................ 20
Hug Transfer ........................................................................................ 20
Materials and Method Used for Removing Transferred Fibers .............................. 21
Statistical Methods Used .......................................................................... 22
Chapter 4
Results ................................................................................................ 25
Intra-comparison of Raw Mean Shedding Index ............................................ 25
Fabric Structure and its Effect on Shedding .................................................. 26
A Fiber-Transfer Study ......................................................................... 27
The Hug Transfer ................................................................................ 29
Results of Spearman R Computation ......................................................... 3]
Repeated Hug Contacts ......................................................................... 31
Chapter 5
Conclusions .......................................................................................... 32
The Significance of a Sheddability Index in the Laboratory and the Courtroom.......38
Chapter 6
Future Research ..................................................................................... 43

TABLE OF CONTENTS, continued

Appendices ........................................................................................... 45
Appendix I
Fabric Descriptions ................................................................................. 45

Appendix 11
Fabric Weave Counts .............................................................................. 46

Appendix [II

Summary of Raw Mean Shedding Index Calculations ......................................... 47
Appendix IV
Comparison of Raw Mean Shedding Index between the Same Fabric Types .............. 49

Appendix V
Categorization of Nineteen Fabrics Tested ...................................................... 52

Appendix VI
Results of Hug and Handshake Transfers ........................................................ 53

Appendix VII
Calculation of Mean Transfer ..................................................................... 55

Appendix V111
Results of MINITAB Statistical Program ........................................................ 56

References ............................................................................................ 57

LIST OF TABLES

Table 1
Example of Shedding Data for a Woven and Knit Fabric ..................................... 25
Table 2
Number of Fibers Transferred during Hug(s) and Handshake ................................ 29

vi

LIST OF FIGURES

Figurel

Photograph of Materials Needed for Determining Shedding Potential... ......................... 17
Figure 2

Raw Mean Shedding Index Comparison of Eight Polyester Fabrics Tested ................ 26
Figure 3

Plot of the Affect of Fabric Structure on Shedding ............................................. 27
Figure 4
Data Yielding Five Shedding Categories ......................................................... 28
Figure 5

Ratio of Raw Mean Shedding Index to Mean Hug Transfer ................................. 3O

INTRODUCTION

The purpose of this research was to develop a standard method for determining
the sheddability of a fabric that is efficient, reproducible, and realistic to real-world
conditions. More importantly this research investigated whether there is a useful
relationship between raw mean shedding index and transfer rate. The raw mean shedding
index of a fabric type will assist the examiner in determining the significance of the fiber
evidence whether it is a single fiber or a tuft of many fibers.

The presence or absence of fiber evidence can have considerable influence on
linking a suspect to a crime scene or particular location. As the mass production of
fabrics and garments continues and as we continue to come in contact with each other,
the opportunity for transfer increases. Understanding how and why fibers are present in
criminal and civil circumstances is essential for a scientist to analyze and interpret the
probative value of the evidence at hand. Criminal investigators and forensic examiners
should not underestimate the evidential value of fibers. Fiber evidence speaks volumes to
the environment, surroundings, and circumstances that a person or object may have been
in at a given time.

Edmond Locard was the first to recognize the significance of evidence transfer.
Locard established what is now known as his exchange principle [10]: contact between
two objects always results in transfer of some material. This remains a central principle
of the analysis of scientific evidence. When applying this principle to fiber evidence,
three components must be considered: shedding, transfer and persistence.

Fiber transfer has been generally defined as the release or shedding of a fiber from

one garment or textile onto either another textile or non-fiber material. Fiber shedding

can be defined as a type of transfer. Research has shown that the dominant method of
transfer is mechanical [18]. Factors involved in how easily a fiber transfers include the
length; cross-sectional shape; weave type of the donor garment; age and environmental
effects; the force and duration of the contact; and the number of contacts; as well as the
characteristics of the recipient garment [l3]. Logically, one would expect short loose
fine fibers from older garments to transfer more readily than fibers from tightly woven
newer garments.

Pressure and contact are the main components necessary for a fiber transfer to
occur [14]. However, how the contact and pressure are involved in a fiber transfer may
vary. For example, an individual may be wearing a highly sheddable shirt made of loose,
acrylic fibers. This individual’s shirt could shed fibers simply from the mechanical
forces the individual exerts on the shirt when wearing it. Without any contact, many of
these loose fibers could then be transferred to another individual nearby who is wearing a
brushed denim shirt-if the two individuals brushed shoulders, for example. If these same
two individuals were to hug one another the chances of the loose acrylic fibers being
transferred to the brushed denim shirt increase, as does the number of fibers transferred.
Therefore, fiber transfer can occur in either situation (direct contact or no direct contact)
when two individuals are in the same environment. The number of fibers transferred in
either example would probably be relatively high because the donor garment was made
from loose acrylic fibers and the recipient garment was made from brushed denim, which
is a good recipient for those loose acrylic fibers.

A fiber transfer between two garments or textile objects in which a minimal

amount of fibers are transferred can also provide information for instance, if the same

two individuals in the former example are now wearing a tightly knit polyester shirt
(donor) and a nylon jacket (recipient). The tightly knit polyester shirt will not shed fibers
as easily as the loosely constructed acrylic shirt. In addition, the nylon jacket has a
smooth surface that does not retain shed fibers as well as brushed denim. The
significance of finding a single polyester fiber on the nylon jacket may be as informative
as finding a cluster of acrylic fibers on the brushed denim shirt.

Persistence is the determining factor on whether or not such items as fibers may
be found on a person or object after a transfer. Persistence affects the chance of finding a
foreign fiber on a recipient after any interval of time has passed since initial transfer.
Persistence is affected by such variables as time, activity of wearer, washing, location of
transferred fibers, and type of recipient material. Although there are many factors to
consider after a transfer has occurred, the shedding potential of a garment must be
considered when evaluating both the transfer and persistence of fibers. A loss of fibers
after transfer should be expected and this makes interpreting the results of a fiber analysis
problematic. If the shedding potential of a textile could be determined, however, the
interpretation of the fiber(s) might be clearer.

It is important for fiber examiners to compare the shedding ability of one fabric
compared to another. Accurately determining the shedding potential of a garment could
provide valuable insight into interpreting fiber evidence that may consist of a large
number of fibers or only a few fibers. Determining the significance of one fiber can be
challenging. There are many variables that affect fiber transfer and having a consistent
method for measuring the shedding potential of a garment would be helpful in

interpreting fiber evidence. A standard method for determining the shedding potential of

a textile could lay the foundation for a uniform interpretation of fiber evidence in forensic
casework.

Previous research by Coxen, Grieve, and Dunlop [2] has noted the inaccuracy of
using Scotch Magic tape for determining the shedding potential of fibers. Therefore, the
low-adhesive backing of a 3”x 3” 3M© post—it note was utilized in this research to
simulate a ‘natural’ shedding of the fabric. Fabric structures for the tested fabrics
included knits and woven fabrics. The fabrics used for this research consisted of a single
fiber type (100% content) because forensic examiners always determine the relative
amounts of foreign fibers found on a recipient garment by counting sole content fibers
regardless of whether it is a fabric blend. Shedding indices for thirty—one different types
of fabrics were assigned after four trials were carried out. A Raw Mean Shedding Index
(RMSI) value was calculated for all thirty—one fabrics. A fiber—transfer study with the
tested fabrics as donors and white/black felt as recipients was performed and compared to
the previously determined shedding index.

The transfer study was conducted as a means for evaluating the raw mean
shedding index data. The classic weighted cigar box transfer method could have been
used, as well as many other published transfer methods. However, the author felt that a
transfer method that reflected human contact could evaluate a fabrics’shedding potential
in a more realistic manner. The purpose of the research was to develop a method for
determining a fabrics shedding potential; it was important not only to develop a novel
method but to also evaluate and assess the usefulness and significance of that method.

With the goal of reflecting human contact, two simulated contacts, a handshake

(no contact), and a hug (friendly contact) were used as transfer conditions. To investigate

whether the number of transferred fibers decreased after a repeated number of contacts,
each hug transfer was repeated three times with each of the tested fabrics using a
different piece of recipient felt. Performing the hug transfer three times also provided a
calculated Mean Hug transfer that could be compared to the Raw Mean Shedding Index
(RMSI). Comparing shedding data from the post-it note method, a type of transfer, and
the hug, another type of transfer, was an important aspect of evaluating the developed
post-it note method.

Utilizing the post-it note method for measuring the shedding potential of a
garment would be helpful in interpreting fiber evidence in several different types of fiber
cases. For example, a fiber examiner receives evidence from a homicide case. The first
item of evidence is a blue acrylic sweatshirt worn by the suspect shooter. The second
item of evidence is a black cotton/polyester t-shirt worn by the suspect upon arrest. The
examiner uses the post-it note method to determine the shedding potential of the acrylic
sweatshirt. Based on several post-it note counts, the examiner determines that the Raw
Mean Shedding Index for the sweatshirt is forty—eight (48). This value indicates that the
sweatshirt is a high—medium shedder. With this knowledge, the examiner can evaluate
the number of fibers transferred on to the suspect’s shirt and determine with assistance of
other chemical tests whether the suspect was wearing the blue acrylic sweatshirt at the
time of the shooting.

The t—shirt is processed for trace evidence by utilizing the taping method and the
scraping method. A total of twenty-five blue acrylic fibers are located in the trace
evidence collected from the suspect’s t-shirt. This finding indicates that twenty-five

fibers were transferred onto the t-shirt from the suspect wearing the sweatshirt and

persisted on the t—shirt from the time the suspect took off the sweatshirt to the time they
were arrested.

By using the post-it note method and determining that the acrylic sweatshirt was a
high-medium shedder, the fiber examiner in the above hypothetical case would be able to
evaluate the significance of finding twenty-five blue acrylic fibers. The acrylic
sweatshirt was predicted to shed a larger amount of fibers. Even though the original
amount of transferred fibers is unknown, the shedding potential of the garment assists in

predicting and evaluating a fiber transfer.

LITERATURE REVIEW
Proof of the principle of transfer for fibers became available in 1975 when Pounds
and Smalldon [13-15] examined the transfer and persistence of wool and acrylic fibers to
various recipient garments. The donor garments were pre-treated with a fluorescent dye
except for the acrylic garment, which already contained native fluorescent fibers.

Treating the garments with fluorescent dye enabled for easier recovery of transferred

fibers. [18]

Transfer and Persistence

Pounds and Smalldon used a contact method that is very common now in fiber
transfer studies. The donor materials were attached to a block and then were pulled, with
a constant pressure over a constant distance, across the recipient garment that lay on a flat
surface. Their experiments also included repeating the numberof contact passes across
the recipient garment for up to eight consecutive contact passes. The fibers were
removed manually or by gentle application of adhesive tape between contact passes.

From their experiments Pounds and Smalldon [13-14] determined that the transfer
and persistence of fibers are significantly affected by pressure of contact, the nature of
the recipient garment, the number of repeated contact passes, and fiber length.

The experimental data obtained from the work of Pounds and Smalldon provided
the scientific community with a solid foundation for future fiber research. Kidd and
Robertson [7] researched the transfer of wool, acrylic, cotton, and polyester/viscose fibers

to and from clothing typically encountered in criminal cases. The garments chosen for

this study were composed of either brightly colored fibers or they possessed inherent
fluorescence to allow for easy assessment of fiber transfer.

Not only did the authors use a wider range of textile fabrics in their study but they
also used a contact method that included hand and machine applied pressure. The
pressure of contact was varied and contact passes were repeated.

Kidd and Robertson confirmed the major conclusions reached by Pounds and
Smalldon. The number of fibers transferred depends on the nature of the recipient fabric
with respect to the recipients’ texture. However, when cotton and polyester/viscose
donor fabrics were used in the study, the nature of the donor was found to be significant
also. The authors compared their applied pressures to a simple experiment with hand
pressure and found that there was an obvious increase in the number of fibers being
transferred. This concurred with Pounds and Smalldon; however, Kidd and Robertson
found that when a maximum or threshold pressure was reached, no additional fibers
would transfer. They also agreed with the previous study, that with repeated contacts the
number of transferred fibers decreases.

Pounds and Smalldon [13] also investigated the persistence of wool and acrylic
fibers on various articles of clothing during various times of wear. Their experiments
showed that the time of wear was the most important factor in fiber persistence.
According to the authors, after an initial rapid loss of fibers there follows a lower loss
with a constant low level of fiber persistence extending after time of wear. The rate of
fiber decay was the same for the wool and acrylic fibers even though garments differed
considerably in surface texture. However, Pounds and Smalldon found that fibers were

lost at a significantly higher rate from smooth garments like a cotton laboratory coat.

Kidd and Robertson [8] conducted a follow-up study on the persistence of fibers
transferred from acrylic, wool and polyester/viscose garments to a variety of recipient
garments. They found that fiber persistence was lowered by several factors including: the
length of wear, low pressure during initial transfer, and fiber size.

Although Kidd and Robertson had used a range of donor and recipient fabric
types, the results of their study confirmed the conclusions of Pounds and Smalldon.
Their research also presented the importance of other factors on the rate of fiber loss,
such as the position of contact and wearing outer garments.

Research on fiber shedding was also conducted using blended fabrics. Parybyk
and Lokan [12] and Salter et al. [17] conducted similar studies, using blended fabrics as
donors. The study conducted by Parybyk and Lokan [12] focused on the difficulties in
predicting proportion of transferred fibers based on the manufacturer’s composition label.
The fiber composition reported on a manufacturer’s label is calculated as percent by
weight. However, the percent by weight is not an accurate indicator of the numbers of
shed fibers. The author’s hypothesis was to determine if the proportion of shed fibers on
a garment could be predicted by knowing the composition of the original donor garment
in terms of the relative numbers of each component fiber. Four bicomponent donor
fabrics were used for their research. The simulated transfer was performed with a block
wrapped in nylon that was pulled across the donor fabrics. Parybyk and Lokan
determined the theoretical and actual ratios of fibers and found that there was good
agreement between the theoretical and actual transferred fiber ratios. However, the
authors found differences between the percent composition by weight of the fibers and

the actual number of fibers transferred. When the theoretical fiber transfer ratio for a

garment and the actual transfer ratio differed, the authors noticed that this occurred in
garments containing wool. Their observation was consistent with previous research that
wool-containing fabrics shed wool to a greater extent than other fibers.

Salter, Cook and Jackson [17] performed a similar study with polyester/wool,
polyester/viscose and polyester/cotton blends. Cotton and polyester garments were used
as recipients during the simulated transfer. Salter et a]. conclusions supported Parybyk
and Lokan that the weakest fiber in the mixture (wool and cotton) always shed more
fibers and that the percentage content of a fabric does not reflect actual shedding. Salter
et al. also found that a rougher recipient material increased the number of fibers
transferred from the minor component of the donor blend. Contact pressure also had an
effect on the proportion of fibers shed from the minor component. Another related aspect
of this research was that the authors used Sellotape (cellophane adhesive tape) as a rough

guide to determining which component shed most from a blend.

Secondary Transfer

The goal of fiber transfer research initially was to assess the factors involved in
primary transfer and persistence. Not until Pounds and Smalldon and other researchers
built a foundation for fiber transfer was research conducted on secondary and higher
order transfers. Grieve, Dunlop and Haddock [4] conducted a classic study on secondary
transfer that was inspired by an actual homicide case. The authors recognized that
forensic scientists often have difficulties interpreting casework findings because the
original number of fibers transferred during an offense is unknown, thus making it

difficult to determine the significance of fiber evidence. The authors criticized previous

10

studies for their lack of realism specifically in regards to the size of the donor area being
too small and contacts being artificial. According to the authors, extrapolating results in
an attempt to fit real life conditions makes the invalid assumption that transfer will be
made with equal pressure on all areas of the recipient and that fibers will be equally
transferred in every unit area [4].

Due to the nature of a homicide case the authors were working on, they set out to
conduct a number of transfer experiments using a red acrylic sweater belonging to the
suspect in the case. They investigated primary and secondary transfer and persistence of
these fibers onto clothing and seats. The experimental condition of a 15-second struggle
kept the number of transferred fibers to a countable level. The authors found that even
after a short lS-second struggle, the number of fibers transferred exceeded those
transferred by shaking or laying the sweater on a recipient item. They also confirmed
previous observations about secondary transfer from clothing and noted that a donor
garment such as a high shedding acrylic sweater could transfer a large number of fibers
onto a garment worn under the donor, i.e. a shirt. In addition, the authors found that
primary-transfer fibers were shown to persist on various seats even after multiple

secondary contacts.

Fabric Sheddability

Coxen, Grieve, and Dunlop [2] researched a method for assessing fabric shedding. The
authors’ goal was to develop a method that was easy to execute, reproducible and
applicable to real life conditions. Their research design was two-fold. To assess the

shedding potential of a fabric, the authors placed a 7.5 x 2.5 cm piece of Scotch Magic

tape on the fabric with applied hand pressure. Ten random tapings were taken and a
score, ranging from I (very poor shedder) to 5 (very high shedder) was assigned based on
the number of target fibers removed.

Coxen, Grieve, and Dunlop [2] also conducted a transfer study as a
complimentary part to their fabric shedding research. The authors believed that the
ability to estimate a garment’s shedding potential would help determine whether that
garment was a potential donor of “target” fibers and whether recovered fibers were the
product of a primary versus a secondary transfer. The donor garments chosen for their
transfer experiments contained only one fiber type. The recipient items were selected to
reflect the types of exhibits typically processed in their laboratory. The method used for
the transfer experiments utilized a weighted cigar box. The donor garment was placed
flat on a bench and a weighted cigar box with a square of the recipient fabric attached to
it, was drawn by hand over the donor garment surface. Time,’weight, distance and
pressure were fixed for the three transfer trials carried out. Fibers were lifted off
recipients by using Scotch Magic 810 tape. The fibers were scanned and counted using a
grid under a stereomicroscope.

They found that wool donors transfer a high number of fibers but that the taping
method tended to over-estimate and under-estimate, depending on the garment, the
shedding potential. Cotton donors were influenced by the nature of the recipient fabric
but the tape method generally over—estimated the shedding capacity. The acrylic donors
tended to be low shedders and the tape method over-estimated their shedding potential.
However, the authors found that the taping method gave accurate assessment of shedding

ability for the polyester donors.

12

The authors recognized that the method they developed for assessing the shedding
potential of fabric does not have practical applications on a routine basis. Coxen, Grieve,
and Dunlop [2] concluded their research by suggesting that recourse to their method

could yield a more accurate indication of the ability of a fabric to shed fibers.

Hypothesis of Current Study

The goal of the current research was to develop a method that would give a more
accurate indication of a fabric’s shedding potential. The hypothesis to be tested in this
research is whether there is a useful relationship between raw mean shedding index and
transfer rate. Using the same criteria as Coxen, Grieve, and Dunlop [2], the author
intended to develop a method of determining fabric sheddability that was efficient,
reproducible, and realistic under real-world conditions. There are several differences
between the previous research and the current study. The use of a 3M post-it note as a
less aggressive means for determining shedding is most notable. The use of a 2-pound
weight, instead of hand pressure, to the applied post—it note makes this method less
variable and therefore more reproducible. The current method yields a numeric value of
shedding potential instead of categorizing shedding from “poor shedder” to “high
shedder.” In addition, one type of recipient fabric was used in the transfer study to
control the affect that recipient material has on shedding.

A novel method was used for the transfer study in the current research. The
forces experienced during contact between individuals during a primary transfer are

difficult to predict. However, comparing results from a weighted cigar box transfer to

those experienced in real life is inaccurate. Thus, this research intended to use a transfer

method that reflected a common interaction, a hug, between two individuals.

14

METHODS AND MATERIALS
To provide fiber examiners with an efficient, reproducible and reliable method for
determining the sheddability of various fiber types, all materials used in this research
were chosen to be inexpensive and commercially obtainable. The laboratory equipment

used in this research is likely to be available in most forensic laboratories.

Fabrics Chosen for this Research

Thirty fabrics were bought at a JO—ANNTM Fabrics and Crafts® store for this
study. One fabric, the acrylic, was donated and is manufactured by KIN KA FIBER CO.,
LTD., Taiwan. All thirty—one fabrics used were sole content fabrics. The exact
percentages of fiber content for each fabric were not checked. However, in an attempt to
verify that each fabric was composed of one type of fiber, fabrics were mostly bought
from fabric bolts. Fabric bolts list the material description. SOme remnant fabrics were
bought and also listed the fabric content. The following types of fabrics were obtainable
at the authors’ local JO—ANNTM fabric store: cotton, polyester, rayon, wool, and nylon.
The fabric samples were chosen to represent the variety of fabric types used in different
garments. Twenty-four woven and seven knit fabrics were chosen. A full description of
all thirty-one fabrics is in Appendix I.

Warp and weft counts were determined for eighteen woven fabrics. A warp yarn
is defined as the yarn which runs parallel to the selvage or longer dimension of a
fabric.[5] The weft yarn runs perpendicular to the selvage or longer dimension of a

fabric.[5] Twelve of those fabrics bought had the selvage end which is the long, finished

15

end of a fabric bolt. This provided a true fabric count. Appendix II outlines these
eighteen fabrics.

After purchase, fabrics were not washed, scraped, taped or vacuumed but used “as
is”. The decision not to wash the thirty—one fabrics was based on the results of previous
research by Colyer [l]. The research conducted by Colyer [1] tested the affect that age-
related wear has on fiber shedding and subsequent transfer. By repeatedly washing
several garments, Colyer found that garments shed less through time but do not do so in a
consistent manner. Washing the thirty-one fabrics used in the current research is outside
the scope of this project. The fabrics could have been washed and then tested using the
post-it note method developed but this would have required washing each fabric piece
thirty-one times using separate washing machines to eliminate contamination. In

addition, each machine would have to be clean prior to subsequent washings.

Materials Utilized for Determining Shedding Potential

After purchase, the fabrics used in this study were cut into 12 x 12 inch swatches.
These swatches were then divided into four, 6 x 6 inch squares using a ruler and a pencil.
These large fabric swatches were packaged separately in plastic bags. Several packages
of 100 sheet, 3 inch x 3 inch yellow 3M Post-it© Notes ([product number 654RP]) were
purchased. One package of a 4 block, 50 sheet, variety colored pack of 2 7/8 inch x 2 7/8
inch 3M Post-it® Note was also purchased. The blue post-it note from this package was
used on lighter colored fabrics for contrast. Figure 1 is an example of the materials

utilized for determining the shedding potential of a fabric.

16

’lgtlul'mi'ljll’lllllll'ljllllillll‘llli13!]1'13!”lll|ll||l'llljll|l|l|l‘l|:llllllllj'lIIIIIIIIu

'- ,
';‘-’j (Ch 312 .223 9330 fiflfifltfim

wameuSmlcu
9: rl r ' n o n r r t ,2 !
IllllillllIflllill'lIIIll’llliIIlllIllllilillI:ll‘rllllIl”all!IIllllIllllIlMlIIHlIlllfillllltlllsllllllllllIllllIllIlilill’zllllilIll'lllllllll!m

 

Figure 1: Example of Materials Needed for Determining Shedding Potential
Two boxes of Xerox 8 1/2 x l 1 clear laser/copier transparencies were also
purchased. A grid of five lines per inch was photocopied onto each of the transparencies.
A two-pound weight is needed for the method developed in this study. For this
study, the two—pound weight used was a heating block for PCR test tubes. Any object,
which sufficiently covers the width of the post-it note, with a weight of exactly two
pounds, will work. In addition to the above materials, a stopwatch or nearby clock is

needed for this method.

Method Utilized for Determining Shedding Potential
With the aforementioned materials, the method developed for determining the
shedding potential of fabrics is simple. As with any type of fiber examination, a clean

working surface is crucial as well as a clean room. The ambient environmental effects of

the laboratory were not controlled for while testing the post—it note method and therefore
the resulting affect such factors as humidity had on shedding cannot be determined.

The method used for determining the shedding potential of the thirty-one fabrics
is as follows. A 12 x 12-inch piece of fabric, divided into 6 x 6 inch squares, was placed
on a clean working surface finish side up. A post-it note was peeled from the top of the
pack and placed adhesive side down onto an area within the 6 x 6 inch square. The two-
pound weight was immediately placed on top of the post-it note, paying special attention
that the weight covered the entire area where the adhesive was. The stopwatch was
started and allowed to run for 5 seconds. After 5 seconds, the weight was removed and
the post-it note was lifted off the fabric and placed, adhesive side down, directly on to the
gridded transparency. The back of the post-it note was used to label the type of fabric
being tested and the area, (i.e. upper left). The process was repeated in each of the four, 6
x 6 inch areas on the fabric. The procedure of post-it noting a fabric and transferring it to
a transparency took roughly 13 seconds. Multiple post-it notes were placed on a
transparency sheet by cutting the non-adhesive side of the post-it note off to allow for
more space on the sheet. For pemranent storage, the edges of the post-it notes were taped
down to the transparency using scotch tape and then the transparencies were stored in
manila envelopes.

The only laboratory instrument needed to obtain the numeric shedding value for a
fabric using the above method is a compound microscope. In this study, fibers adhering
to the adhesive side of the post-it note were counted using a Fischer Scientific compound
microscope at a magnification of 45. Oblique, transmitted and reflected light were used.

Both target and non-target fibers were counted separately. Target fibers were defined on

18

the basis of color. Varying light sources were utilized to aid in distinguishing between
target and non—target fibers. Blue post-it notes were used to provide contrast with the
light colored fabrics. The post-it note method was repeated four times for each fabric to
obtain four measurements of shedding. An average of these four numeric counts was

calculated to assign a Raw Mean Shedding Index (RMSI).

Materials Utilized for Transfer Studies

Fiber—transfer studies were performed with nineteen of the thirty-one fabrics
tested using the previous method. The materials needed to perform this fiber-transfer
study are easily obtainable and of minimal cost. The nineteen fabrics used as donor
fabrics in the fiber transfer are marked with an asterisk in Appendix I. The recipient
fabric chosen for this transfer study was white and black felt. Approximately ten yards of
felt were purchased from a JO—ANNTM Fabrics and Crafts® store (nine were white and
one black). After purchase the felt was cut into 12 x 12 inch swatches. The laboratory
attire required for this transfer study included a tightly woven t-shirt and a laboratory
coat. Participants in this study always wore light—colored t-shirts and disposable
laboratory jackets buttoned to the collar during the procedures.

Additional materials included boxes of large plastic storage bags, four large safety

pins, and an ink pen.

Method Utilized for Transfer Studies
For this fiber-transfer study two participants were needed. The author recruited a

fellow graduate student to assist her in this study. Both participants wore a light colored

t-shirt and a laboratory coat buttoned to the collar. Since the laboratory being used for
this research did not have a designated “clean room”, a room void of other fabrics or
textiles was used to prevent transfer from items other than the fabrics being tested. Each
participant safety pinned the fabric, either donor fabric or recipient felt, to the front chest
area of their laboratory coat. A large plastic storage bag and an ink pen were always

nearby.

Handshake Transfer

After securely fastening a piece of fabric to our laboratory coats, we proceeded to
shake one another’s hand. To develop a reproducible handshake each time, we stepped
towards each other and shook hands. Thus, the handshake method was: step forward,
shake hands, release handshake and step back. The participant wearing the donor fabric
would remove the recipient felt from the other participant and fold it twice over onto
itself preventing the loss of any transferred fibers. The folded recipient felt was placed
inside a plastic storage bag and labeled. This procedure was repeated with a new piece of

recipient felt and a new donor fabric.

Hug Transfer

To develop a reproducible hug each time, the same two individuals performed all
the hug transfers. Having the same two participants kept pressure relatively consistent.
The method for the transfers via hug was as follows: step forward towards the other
individual, embrace in a non-violent hug, release hug and step away. The participant

wearing the donor fabric would remove the recipient felt from the other participant and

20

fold it up and then over onto itself preventing the loss of any transferred fibers. For each
of the nineteen donor fabrics, the hug transfer method was repeated three times per donor
fabric, using a new piece of recipient felt for each transfer. The folded recipient felts,

were placed inside separate plastic storage bags and labeled Hug 1, Hug 2, or Hug 3.

Materials and Method Used for Removing Transferred Fibers

After the transfer study was completed, the recipient felts were examined for
transferred fibers. The most efficient and common way of recovering transferred fibers is
the tape method. For this research, the adhesive strip (5” x 7”) of a Lynn Peavey tape
lifting kit was used. To conserve tape, the adhesive strips were cut in half horizontally.

Each recipient felt was examined separately by removing the folded felt from the
plastic storage bag and placing it on a clean working surface. The fabric was then
carefully unfolded. The taping procedure consisted of removing the backing of the
adhesive strip and placing the adhesive strip, adhesive side down, on the felt and rubbing
the back surface of the tape with the forefinger. This process was repeated in a
systematic manner until the entire piece of felt was taped. Adhesive strips were
preserved by placing them adhesive side down on to a gridded transparency, which was
then appropriately labeled.

In this study, fibers adhering to the adhesive strip were counted under a

Fischer Scientific compound microscope at 45X. Oblique, transmitted and reflected light
were used. Both target and non-target fibers were counted separately. Target fibers were
again defined on the basis of color. Fibers were counted and reported for the handshake

transfers and for the hug transfers. Each donor fabric was subjected to three trials of the

21

hug transfer thus resulting in three tapings, from three pieces of recipient felt, for each
hug. Labeling was important in this part of the study to maintain sample integrity.
Repeating the hug transfer led to three “Hug” measurements. A mean of these
measurements was calculated and for each of the nineteen fabrics tested a “Mean Hug

Transfer” value was assigned.

Statistical Methods Used

If there were a perfect linear relationship between the two variables Raw Mean
Shedding Index (RMSI) and the Transfer Mean, then for every fiber shed, one fiber
would be transferred. This assumption will only be true if in the population the variable
of interest (fabric sheddability) is normally distributed. Current and previous research
has established that there are many variables involved in fabric shedding and fiber
transfer. Due to these variables, fiber shedding is not normally distributed. Therefore,
application of classic parametric statistics is not applicable to this research data.

Another factor that limits the use of parametric tests to this research is the size of
the sample data to be analyzed. Sampling distribution may be assumed normal, even if
there is uncertainty whether the variable in the population is normal, as long as sample
size is large (n 2 100). Only nineteen (n=19) out of the thirty-one fabrics were subjected
to the hug transfer.

To evaluate the data from this research a need for statistical procedures that allow
for processing low quality data. from a small sample, on variables that are not normally
distributed are needed. The application of nonparametric methods is therefore utilized

for this research. When conducting the current research the calculated mean and standard

deviation for the shedding index were not intended to be applied to a Gaussian
distribution. The repeated measurements taken in this research were for internal
consistency and for the ability to calculate a mean that was representative of the sample.

To express a relationship between two variables one usually calculates the
correlation coefficient. There is not a perfect correlation between Raw Mean Shedding
Index (RMSI) and Hug transfer so the nonparametric equivalent to the standard
correlation coefficient was needed. There are three commonly used types of
nonparametric correlation coefficients. These include, Spearman R, Kendall Tau, and
Gamma coefficients.

Spearman R “assumes that the variables under consideration were measured on at
least an ordinal scale, that is, that the individual observations can be ranked into two
ordered series.” [9] Spearman R would be an appropriate test for this research data. The
individual measurements of RMSI and Hug transfer can be ranked into two ordered series
because intervals between numbers on an ordinal scale are not necessarily equal and the
higher numbers of fibers shed or fibers transferred represent higher values on a ranked
scale.

Kendall Tau is similar to Spearman R but instead of computing from ranks,
Kendall Tau “represents a probability that in the observed data the two variables are in
the same order versus the probability that the two variables are in different orders.” [9]
The hypothesis for this research questions the relationship between the variables RMSI
and Hug transfer, not the probability that the two variables are different. Kendall Tau is

not the appropriate method to apply.

23

Gamma is also a probability like Kendall Tau, but “it is computed as the
difference between the probability that the rank ordering of two variables agree minus the
probability that they disagree, divided by 1 minus the probability of ties.” [9] The gamma
method is not applicable to this research because the data is too unpredictable to
determine whether the factors affecting shedding and transfer are identical for the same

fabric at the same time (probability of a tie).

24

RESULTS
Each quadrant of a fabric was tested using the post-it note method and the
corresponding target fibers were counted. The numeric values for these counts are
designated as A, B, C, and D in Table 1. These values are in no specific order. A
summary of the results for all thirty—one fabrics can be found in Appendix III. The
numeric shedding value for the thirty-one fabrics used in this study was obtained by
taking the average value of the four numeric counts. This number is reported as the mean

shedding index. Standard deviation calculations were also performed and are reported.

 

 

 

 

 

 

 

 

 

 

 

Woven Fabric A B C D Mean Shedding Index Std. Dev.
[Poly. Hot Pink Fleece 64 17 6 8 24 27
[Non-Native Fibers 11 11 10 8 1O 1
Knit Fabric A B C D Mean Shedding Index Std. Dev.
Cotton-maroon knit 343 264 328 444 345 75
Non-Native Fibers 19 6 31 3 15 13

 

 

 

 

 

 

 

 

 

 

Table 1: Example of Shedding Data for a Woven and Knit Fabric

Intra-comparison of Raw Mean Shedding Index

For this research, 8 polyester fabrics, 3 rayon fabrics, 10 cotton fabrics, 6 wool
fabrics, and 3 nylon fabrics were tested. The raw mean shedding indices for each type of
fabric were compared to investigate the differences in shedding between different fabrics
of the same type. For example, the polyester fabrics tested in this research were both knit
and woven fabrics. The fabrics varied in color and print effect. Some fabrics were
intended to appear soft like a fleece material while others were tightly knitted for a
smooth finished look. These differences contribute to the variation in shedding index

among the same type of fabrics.

25

A Graphical representation of the intra—comparison of the polyester fabrics is shown in

Figure 2. The graphs for the additional fabric types can be found in Appendix IV.

F————’ .7017 * _ n," ’_f ,‘___._._v.___._— ’—“—u _.*.__._7 f i "a. 7 .i

Polyester Fabrics l

300 —
250 ~
200 ~-
150 a
100 a

RMSI

 

 

r-ltanlr—fij—r-
1 2 3 4 5 6 7 8
Fabrics Tested

 

Figure 2: Raw Mean Shedding Index Comparison of Eight Polyester Fabrics Tested

Fabric Structure and it’s Effect on Shedding

After comparing the raw mean shedding index between fabric types it was
relevant to investigate whether fabric structure had an affect on shedding. The woven
fabrics chosen for this study included plain weaves, twills, satin, crepe, and a dobby
fabric. There was an insufficient amount of knit fabrics obtained for this research. Due
to this, no conclusions could be made on whether fabric structure has an effect on
shedding. However, the graphical representation of this comparison warrants inclusion.

(Figure 3)

26

500 ~ 1
450 J
400 i
350 ]

300 4 . .
250_ O IWOVENS l l

l

 

9 'CKNEeG l l

200] ’ ‘
1501 ‘

Q l
100 .-

-, _ , , f . l .EL . r
O 5 10 15 20 25 30 35
Fabfics

Fibers Shed

 

 

 

 

Figure 3: Plot of the Affect of Fabric Structure on Shedding

A Fiber-Transfer Study

Comparing the Raw Mean Shedding Index (RMSI) data of the fabrics’
demonstrated differences in the sheddability of the fabrics. The raw mean data itself
divided fabrics into the following shedding categories: < 30 fibers shed, 30-60 fibers
shed, and > 60 fibers shed. However, after calculating the medians for each fabric and
plotting that data, natural breaks suggested five distinct categories. A light gray line at
the mid—point between the adjacent values designates these “natural breaks” in Figure 4.
This broke the data up into more categories and seems to fit the data better. Fabrics were

categorized based on this division.

27

150
140 ~

130 «
120 1 DataThreshold

 

110~ -

100 T

 

803

70

Fibers Shed

60

50~

 

4o

30 _

20 j H
10 - -'""_"

 

 

Fabrics

Figure 4: Data Yielding Five Shedding Categories

Fabrics were categorized as the following: high shedders (X>80), high-medium
shedders (80>X>46), medium shedders (45>X>26), medium-low shedders (25>X>l6),
and low shedders (X>15) with X being the mean shedding index calculated after using
the post-it note method. Mean shedding indices are reported in Appendix III. A
representative sample of each fabric type and shedding category was chosen for the fiber
transfer study. The categorization of fabrics used for the fiber transfer study is presented
in Appendix V. A total of nineteen out of thirty-one tested fabrics, were used as donors

for the fiber-transfer study.

28

The fiber transfer study itself involved two types of casual everyday contact, a
handshake and a hug. Table 2 is an example of the number of fibers transferred during
these simulated contacts. The hug transfer was repeated three times for each fabric. A

summary of the results for the nineteen donor fabrics can be found in Appendix VI.

 

 

 

 

 

 

 

 

 

 

 

 

Number of fibers transferred: HUG 1 HUG 2 HUG 3 HANDSHAKE
Acrylic 170 165 83 NR
Non-Native Fibers 3O 18 12
Poly.Black Suede-knit 757 576 456 3
Non-Native Fibers 113 101 94 5

 

 

Table 2: Number of Fibers Transferred during Hug(s) and Handshake

The Hug Transfer

No significant data was obtained from the handshake transfer. A mean was
calculated from the data of the three hug transfers. The calculated means can be found in
Appendix VII. The purpose of the transfer study was to answer whether there is a useful
relationship between Raw Mean Shedding Index (RMSI) and transfer rate. The data for
each was compared by plotting the ratio of RMSI to mean transfer for all nineteen-donor
fabrics. Figure 5 exhibits this graphical data. The ratio comparison shows that fibers are
being transferred from the fabrics more than shed over 50% of the time. This was
expected since the pressure exerted during the hug transfer is higher than the pressure

exerted by a two-pound weight.

29

 

 

 

 

 

 

 

 

 

 

 

 

50%

 

 

1.40 *
O
_ _-____._-_Q;_-z A E _ _,, __--_# _
120 I‘ . Over-Estimates
’ o
. . 9
0.80 e ~ -,u._zzz--__¢____,_____ — —
O . Under-Estimates
a s .
“oeoeee— -
O
0 o
0.40 -._-.g___9 ee —-—-———
O O. O Q 0
O
0.20 -2”- , mflz+ 2‘2
. O
0.00— o ’ . H e
O. 20 Fabrics 4O 60
Under-Estimates by more than

Figure 5: Ratio of Raw Mean Shedding Index to Mean Hug Transfer

30

Results of Spearman R Computation

Nonparametric statistics are less statistically powerful than parametric methods,
and if the goal is to detect small effects to the data then caution should be taken when
choosing a test statistic. For the research at hand and the sample data collected, a
nonparametric test like Spearman R is sufficient to determine whether there is a positive
correlation between the variables RMSI and Hug transfer. Figure 5 shows that there is
not a perfect linear relationship between RMSI and Hug transfer. However, there is a
relationship between the two variables. The formula for Spearman’s Rank Correlation

Coefficient is:

Rs: ssxy/V(ssxyssxy)
SSxx = Z(x—§)2 . SSW: 2(y-§)2 , ssxy=z(x—i)(y—§)

Using the MINITAB Statistical Software Version 13 @2000, a Spearman R value was
calculated. For this data set of n=19, the correlation coefficient equals 0.753. The data

spreadsheet generated from the MINITAB software is shown in Appendix VH1.

Repeated Hug Contacts

Hug contacts were repeated to also study whether the number of transferred fibers
decreased after repeated number of contacts. The number of fibers did decrease after
multiple contact passes (hugs) for all nineteen fabrics tested. This data is presented in

Appendix VI.

31

CONCLUSIONS

Due to the complex nature of fiber transfer and persistence, it should be
emphasized that the determined shedding index for the fiber types tested in this research
be viewed as preliminary. However, the method developed for determining the shedding
index of fiber types could be useful to a trace evidence examiner. The low adhesive
backing of the post-it note did serve as a less aggressive means for determining shedding.
The post—it note is a common laboratory supply that an analyst could quickly use to test
the sheddability of a suspected source garment. Based on the fiber type of the source
garment, the analyst could then classify the source garment as a high shedder, low
shedder, or somewhere in between, using previously determined sheddability indices like
the one presented in this research.

The method parameters used in this research are reproducible, thus a laboratory
could conduct a similar experiment to determine the shedding indices of common
garments presented in casework. The fabrics used in this research do not completely
reflect the types of garments presented in casework. The author did not have access to
laboratory evidence and was constricted by the fabric stores available in the area.
Garment types will also vary among laboratories based on the climate region that
laboratory is located in. Therefore, it is best if each laboratory repeat the method for
determining sheddability in their respective laboratory environment with the types of
garments typically seen as evidence for that laboratory. If multiple laboratories
participated in such research, cross sharing of data would be useful and could lay a

foundation for a uniform interpretation of fiber evidence in forensic casework.

32

To determine the raw mean shedding index of a garment an examiner would need
to repeat the post-it note method on several areas of the target garment. The post-it notes
could easily be placed on a microscope slide and then screened using transmitted plane-
polarized light for target fibers rather than a stereomicroscope. In this research, color was
the only property used to pick out and count target fibers from the post-it note. In
addition to color, the optical properties of the target fibers should be used to determine an
accurate count for the shedding index.

The transfer study conducted in this research project was unique compared to
previous transfer studies. The classical transfer experiment usually involves a weighted
cigar box with a recipient material attached that is pulled, with a constant pressure over a
constant distance, across a donor garment. Although pressure, time, and distance is all
controlled during these transfers, little can be extrapolated from such artificial contacts.
These fixed conditions make the invalid assumption that transfer will be made by equal
pressure on all areas of a recipient garment. Such conditions do not fit real-life transfers
that occur in forensic casework. Due to the disadvantages found in the classical method,
the goal of the transfer study in this research was to simulate a real life, everyday contact.

When deciding on a method for the transfer study the factors of pressure, time,
and area were considered. A simple gesture such as a hug often occurs between
individuals in day-to-day contact. In order to hug, two individuals must come into
contact. With a hug contact there is a form of pressure based on each individual’s body
weight, strength, and enthusiasm behind hugging the other individual. Although, these
three things may not be the only factors involved in such contact they are good variables

to consider as a guideline for this transfer method. The contact area in a hug gesture is

typically the front chest area of each individual. With all three factors considered, the
hug transfer method used in this research project was developed.

The hug transfer method requires fewer materials than the classic weighted cigar
box. By using the same two individuals as participants in the hug transfer; the pressure
of each contact could be relatively controlled. A rhythm to the gesture develops between
the two participants that provides method reproducibility. The use of white felt as a
recipient gamrent in this project reduced the number of variables involved in the transfer.
Even though you seldom find in forensic casework a transfer between a donor textile and
a white felt, the purpose of using white felt was to eliminate the affect the recipient has
on shedding. The goal of this research was to first develop a reliable method for
determining the shedding potential of a garment, and then to test whether this method
was useful when evaluating fiber transfer. Therefore, the focus was on the donor material
and the amount of fibers that material shed and transferred. The affect of the recipient
material needed to be negligible.

This research project generated two sets of data. The first set of data was the Raw
Mean Shedding Index (RMSI) data that was based on the post-it note method developed.
The second set was the mean Hug transfer that was calculated based on the multiple hug
contacts performed with each fabric tested. As previously discussed, the RMSI data is a
promising tool for a trace evidence examiner to use when evaluating fiber evidence. The
data from the transfer is not meant to be used as a tool in forensic casework. The
examiner cannot determine the variables involved in a transfer. Even with controlled

transfer studies like this one and others in the past, the variables involved are too

unpredictable. The transfer study was conducted as a means for evaluating the raw mean
shedding index data.

When the RMSI data and the mean Hug transfer data were compared by plotting
the ratio of RMSI to mean transfer, the ratio comparison shows that fibers are being
transferred from the fabrics more than shed over 50% of the time. This was expected
since the pressure exerted during the hug transfer is higher than the pressure exerted by a
two—pound weight. However, because such a large percentage of the nineteen fabrics
tested were transferring fibers a significantly larger amount than shedding fibers, pressure
alone is not the only variable involved. Only four of the nineteen fabrics tested actually
shed more fibers on to the post-it note than transferred fibers on to the recipient felt.
These four fabrics included the hot pink polyester knit, the wool twill, denim #1, and the
meshed nylon. The RMSI values for the nylon meshed fabric and the wool twill fabric
were not significantly greater than the mean Hug transfer. However with the denim
fabric, the difference between the RMSI and the mean Hug transfer is sixty-eight fibers.
The difference between the RMSI value and the mean hug value for the hot pink
polyester knit is ten fibers. This value is not as large and it is difficult to conclude
whether either is a significant finding.

There could be many explanations as to why these four fabrics had a higher Raw
Mean Shedding Index. The non-native fibers found on the fabric may have prevented
fiber transfer. The pressure, duration and sequence of the hug method may have been
altered. Fibers may have been misidentified and thus counted, on the post-it notes. The
persistence of fibers transferred on to the white felt after the hug should also be

considered. There is a possibility that fibers could have fallen off the white felt from the

35

time the hug contact took place to the time the felt was folded and packaged away. Any
of these explanations might be true but were not tested for in this research. Repeating
this research using similar fiber types but using optical properties as an additional guide
to sorting target fibers could lead to answers regarding persistence.

The most significant outlier in the data was the wool peach crepe fabric. Based
on the low Raw Mean Shedding Index value (7), this fabric was categorized as a low
shedder. In the transfer study, this fabric behaved like a high to hi gh-medium shedder
having a mean transfer value of 182. Since the shedding potential for only one out of the
nineteen fabrics tested was grossly incorrect, the success rate of using raw mean shedding
indices to categorize the shedding potential of a textile has the potential to be 95%
accurate.

Additional differences in the raw data from the post-it note method and the hug
transfer occurred between the cotton knit fabrics and the denim fabrics. Three out of the
five cotton fabrics chosen for the transfer study were knit type fabrics. The only visible
difference between the three was color. One knit was not dyed while the other two were
maroon and teal in color. The finishing effect on all three fabrics was similar. The
maroon and white cotton knits were categorized as high shedders while the teal cotton
knit was categorized as a high-medium shedder. Although the white and maroon cotton
fabrics were categorized as high shedders, the RMSI and transfer values for each were
quite different. The shedding potential of the teal cotton knit was much lower than either
of the white or maroon cotton knits. A difference in fiber diameter between these three

fabrics may account for the differences in shedding potential.

36

Differences in shedding potential were also evident between the cotton denim
fabrics chosen for the transfer study. The fabric structure for all three denim’s is right-
handed twill. However the fabric construction and weight are different. Denim #1, the
denim with the highest shedding potential, is brushed lighter weight denim. This denim
was probably washed either chemically or mechanically before it was shipped to the
fabric store. The washing loosens the fibers from the weave that in turn causes the fibers
to pill, giving a brushed look to the denim. Loose fibers are easily she'd from a garment,
which could explain why Denim #1 had the highest shedding potential.

Denim #2, the denim with the lowest shedding potential, was stretch denim. Due
to the stretch factor of this denim it is likely that this fabric my have a low percentage
content of spandex and is not 100% cotton. The spandex content could have played a
role in the shedding of the blue cotton fibers.

Denim #3, the denim with a shedding potential between that of the other two
denims, is rough heavy weight denim. This denim does not appear to stretch and has not
been washed. How many other factors such as washing and fabric weight, affects
shedding cannot be determined in this research but based on these results these variables
should be considered when evaluating the shedding potential of a textile.

Although a sound method was developed for determining the shedding potential
of a fabric, the results of the rank correlation Spearman‘s r statistic were valuable to this
research. This nonparametric statistic tested for correlation between the raw mean
shedding index data and the transfer data. Without this test, an evaluation as to whether
the post-it note method for determining shedding could be used to correlate with transfer

evidence could not be made.

The first step in finding Spearman’s rank correlation coefficient (r,) is to rank the

values of each of the variables, RMSI and Mean transfer, separately. The Minitab
software ranked the nineteen data points. The correlation coefficient is then computed in
exactly the same way as the simple correlation coefficient. The only difference is that the
values of x and y are ranks of the raw data rather than the raw data themselves.

The accepted hypothesis (Ha) for this study is that the ranked pairs are positively

correlated. The null hypothesis (H) would be no correlation between the ranked pairs.

The rejection region is r8 2 r0. For this data set of n=l9, the Spearman’s rank correlation

coefficient equals 0.753. Performing a one-tailed test with alpha equaling 0.005, r0 is

equal to 0.608. Therefore, the null hypothesis is rejected.

A positive correlated relationship exists between the raw mean shedding index
and the mean transfer of a fabric. This ultimately means that a fiber examiner could use
the Raw Mean Shedding Index for a fabric type to interpret the evidential value of fiber

evidence in a forensic case.

The Significance of a Sheddability Index in the Laboratory and the Courtroom
The examination and interpretation of fibers in forensic science examinations is
dependent upon a variety of factors. There have been a number of articles published on
the evaluation of fiber evidence and the factors that influence the significance of such
evidence. [18] Whether a factor has been determined to strengthen or weaken the fiber
evidence the possible conclusions (excluding inconclusive) that can arise from a fiber

comparison are still either:

38

a. The unknown fibers are similar to the known in all respects

b. The unknown fibers are not similar to the known
In the courtroom these conclusions are inadequate to the fact finders and should also be
inadequate to the scientist. Fiber examiners are frequently being asked complex
questions regarding the circumstances of the case, the type of transfer, degree of contact
and the pressure of contact. Using Raw Mean Shedding Index data would assist an
examiner in answering these questions. The assumption being that enough scientific
testing is conducted between multiple laboratories, in various locations, where a solid
database is created of the shedding potentials for common casework evidence.

Take for example a case of fiber transfer, either one-way or two-way, where a
victim was assaulted and killed. A suspect is apprehended and fibers are collected from
the suspect’s clothing and environment. The shedding potential of all the clothing
involved is determined. The following two by two matrix illustrates the possible

scenarios a forensic trace examiner is faced with when evaluating the transfer evidence.

High number of fibers transferred Low number of fibers transferred

 

EXPECTED WHY SO FEW?
lHigh Shedder

 

Low Shedder WHY SO MANY? EXPECTED

 

 

 

 

39

In the first scenario, by using the post-it note method, the examiner determines
that the source garment is a high shedder. There is also a large amount of fibers found on
the victim. Since the source garment was determined to be a high shedder, finding a
large amount of foreign fibers on the victim is expected. After analyzing the fibers the
examiner concludes that the suspect fibers are similar to the source garment. In addition
to these analytical results, the examiner can also assess the value of this fiber evidence by
using the shedding index data. Based on the shedding index data, the examiner could
accurately conclude that the presence of so many foreign fibers on the victim is
considered to be indicative of recent contact and little movement of the body after
contact.

Another expected scenario is when the examiner determines that the source
garment is a low shedder and finds only a small amount of fiber transfer evidence on the
victim. After analysis, the suspect garment cannot be excluded as a source of the fibers
found on the victim. The transfer evidence in this case could be more significant since
finding fibers on the victim when the suspect was wearing a garment that was a low
shedder is rare compared to if the source of the transfer evidence had been a high
shedder

In the above two scenarios, shedding index data supports the analytical data as
well as provides further explanation as to the circumstances of the case, the type of
transfer, degree of contact and the pressure of contact. Shedding index data can also help
interpret the unexpected scenarios a forensic trace examiner is faced with when
evaluating evidence. There may be a situation where the source garment has been

determined to be a high shedder based on the post-it note method, but the number of

40

fibers recovered off the victim is low. This scenario does not mean that the post-it note
method has failed to provide valuable information regarding the evidence. There can be
several reasons why there is a lack of fiber transfer from a high shedding source. The
most important thing to consider when evaluating this scenario would be the persistence
of fibers after they have been transferred.

Pounds and Smalldon [13-15] conducted the earliest studies on persistence. They
found that the most significant factor in fiber persistence was the time of wear of the
recipients after transfer of fibers. This could mean that a large number of fibers were
shed and transferred from the high shedding garment but the initial rate of fiber loss was
high due to an extended period from initial contact to victim discovery.

Kidd and Robertson [8] also conducted a study on the persistence of fibers. They
found that fiber persistence was lowered by several factors including: the length of wear,
low pressure during initial transfer, and fiber size. These conclusions would support the
possibility that the contact between suspect and victim was neither lengthy nor forceful.

Lowrie and Jackson [8] conducted a study on the persistence of fibers on a wool
jumper, an acrylic jumper, and a polyester jacket. They found that the combination of the
nature of both donor and recipient had a significant effect on the results of their study.
The authors found that smooth garments had the lowest levels of persistence for all types
of transferred fibers. This study and others like it should be considered when evaluating
the type of garment(s) involved in the transfer.

In addition to the consideration of fiber persistence, another possibility of the high
shedder—low transfer number scenario is that the victim’s body was moved or other

evidentiary items were removed from the primary crime scene. This scenario may also

41

indicate that the fiber evidence was not from a primary transfer but in fact from a
secondary or tertiary transfer.

When faced with this high shedder-low transfer number scenario, an examiner
must also consider that the unknown fibers could be a coincidental match to another
similar source; while highly unlikely given clinical testing in the literature on incidental
matches, it is still possible.

The last scenario a fiber examiner could be faced with is where the source
garment has been determined to be a low shedder based on the post-it note method, but
the number of fibers recovered off the victim is high. Several reasons could account for
this. The source garment could be torn or damaged causing more fibers to shed from the
garment. Careful observation should thus be taken when examining the source garment.
Evidence left at the crime scene of the cause of such damage would also be valuable and
should not be overlooked. The low shedder-high transfer number scenario could also
mean that contact between the victim and suspect may have been lengthy and forceful
resulting in a higher number of fibers shed and transferred. Multiple contacts would also
increase the amount of fibers transferred from a low shedder. Finding a high number of
transferred fibers from a low shedding source could also indicate that a recent transfer
occurred between the suspect and victim.

In the low shedder-high number of fibers transferred scenario, it is more likely
that local effects such as activity, garment damage, and repeated contacts would account
for the transfer evidence more than a coincidental match. The high-fiber result from a
low-shedder means an increased significance of the match. In reality, coincidental

matches happen at a very low rate.

42

FUTURE RESEARCH

Upon conclusion of this study, many additional areas of future research have been
recognized as imperative for assessing the significance of the current findings. One very
important addition to this research would be to conduct the research on a much larger
scale. A larger data set would have to include using more fabrics in all the shedding
categories, using fabrics that are typically found in forensic casework, using more knits,
using more dyed and non dyed fabrics, and using printed fabrics. Conducting a larger
scaled project would help identify all of the variables involved in shedding. Identifying
all variables involved in shedding is crucial when evaluating the significance of a fiber
transfer in forensic casework.

Other research areas to be included in the larger scaled study would be to evaluate
the affect a fibers cross—section and diameter have on shedding. This would entail
gathering textiles of the same fiber type with different cross-sectional shapes and
diameters and intra—comparing the shedding of those as well as comparing those textiles
to the shedding potential of other fiber types.

To the fabrics used in the current study an additional study could be conducted
where the swatches were washed and dry cleaned and re-tested to determine the affect of
washing on shedding index. The significance of a common occurrence such as garment
washing could be evaluated and accounted for when determining the shedding index of a
fabric.

Another area of research interest would be to investigate the persistence of fibers
on the recipient felt. The research question is essentially what would the average fiber

loss be after x, y, and 2 time passed? These results would need to be related back to the

43

original shedding index for that donor fabric and considered when evaluating transfer
evidence.

With all of the above future research suggestions being considered for single fiber
type (100% content) fabrics, using blended fabrics as an additional research variable has
to be considered. Conducting this research study using blended fabrics may or may not
help in determining the shedding potential of a blended fabric. This observation is based
on the results of Parybyk and Lokans [12] study. However, blended fabrics do warrant
consideration in determining shedding potential because they are prevalent in the garment

industry.

44

Fiber Type
*Acrylic

*Polyester
Polyester
*Polyester
Polyester
* Polyester
Polyester
Polyester
*Polyester

Rayon
*Rayon
Rayon

W001
W001
*Wool
*W001
W001
*W001

*Cotton
Cotton

Cotton

*Cotton
*Cotton
*Cotton
*Cotton
*Cotton
*Cotton
*Cotton

*Nylon
Nylon
*Nylon

Appendix I

 

 

Fabric Description Fabric Structure
Brown and came] sweater material Woven
Black, brushed, suede-like appearance Knit

Hot pink fleece material Woven
Blue, silk-like feel as used for pant lining

White, fleece material Plain Woven
Hot pink shirt material Knit

Pink w/orange, fuchsia & purple bubble pattern Crepe Woven
Red w/white polka dot pattern Woven
Navy blue shirt/suit material Knit

Purple, black & pink animal print Plain Woven
Red w/white daisy pattern Plain Woven
Pink shirt material Plain Woven
Tan Even—sided Twill
Black Plain Woven
Tan Twill

Red, black & green plaid Dobby Fabric
Navy blue/black shirt material , Warp-faced Twill
Peach Crepe

White fabric glittery pink snowflake pattern Plain Woven
Bowling shirt-bowling scene pattern Plain Woven
Sage green w/yellow box print pattern Plain Woven
Teal Warp Knit
Maroon Warp Knit
White—Undyed Warp Knit
Blue & white flannel Plain Woven
Denim #1 Right-handed Twill
Denim #2 Right-handed Twill
Denim #3 Right-handed Twill
Light blue mesh material Warp Knit
Greenish-gray rain jacket material Plain Woven
Creme smooth material Plain Woven

*Denotes Donor Fabrics Used in Fiber-Transfer Study

45

Appendix 11

Fabric Description

 

Acrylic-brown and camel sweater
Rayon—red w/white daisy pattern
Rayon-pink shirt material

Rayon— purple, black & pink animal print
Polyester-red w/white polka dot pattern
Polyester-hot pink fleece material
Polyester-white, fleece material
Nylon-creme material
Nylon—greenish-gray rain jacket material
Wool- navy blue/black shirt material
Wool-tan

Wool—black

Cotton-blue & white flannel
Cotton-denim #1

Cotton-denim #2

Cotton-denim #3

Cotton-bowling shirt

Cotton-sage green w/yellow box print pattern

Cotton-white fabric w/effect face glittery pink snowflake pattern

46

Fabric Count

48 x 24

72 x 72

40 x 40

72 warps x 68 fillings
84 warps x 84 fillings
88 warps x 44 fillings
68 warps x 68 fillings
48 x 44

72 x 72

24 warps x 44 fillings
76 warps x 24 fillings
48 warps x 24 fillings
44 x 44

44 warps x 44 fillings
48 warps x 48 fillings
44 warps x 44 fillings
72 x 40

60 warps x 24 fillings

68 warps x 28 fillings

 

Summary of Raw Mean Shedding Index Calculations

Appendix III

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MEAN
FABRIC *A *B *C *D STD. DEV
SHEDDING INDEX

Acrylic 80 44 70 61 64 15
Non-Native Fibers 0 0 0 0 0 0
fl’olyBlack Suede-knit 288 228 351 230 269 58
mortals/e Fibers 21 18 25 12 19 5
[Poly. Hot Pink Fleece 64 17 6 8 24 27
[Non-Native Fibers 11 11 10 8 1o 1
[Poly. Blue Silk-like feel 3 8 7 7 6 2
Mon—Native Fibers 1 0 1 1 1 1
[Poly. Wht. fleece material 9 7 6 8 8 1
[Non-Native Fibers 2 6 2 3 3 2
Ely. Hot Pink knit 52 64 58 54 57 5
[Non-Native Fibers 22 18 30 25 23 8
Ely. Pink w/orng bubble 19 15 23 16 18 4
[Non—Native Fibers 0 0 O O 0 0
“Poly. Red/white polka

dot 19 16 30 18 21 6
Non-Native Fibers 4 14 1O 1 7 6
Poly. Navy blue knit 9 16 16 17 15 4
Non-Native Fibers 6 4 2 8 5 3
Rayon-Animal print 12 12 16 10 1 3 3
Non-Native Fibers O O 0 O 0 0
Rayon-Daisy Pattern 47 72 54 51 56 11
Non-Native Fibers 2 3 6 1 3 2
Rayon-pink 6 2 4 3 4 2
Non-Native Fibers 0 2 2 3 2 1
Tan Wool-evn-sided twill 29 33 24 41 32 7
Non-Native Fibers 26 25 14 15 20 6

 

 

 

 

 

 

 

 

 

47

 

 

 

 

 

 

A endix Ill (cont’d).

 

 

 

 

 

 

 

 

 

 

 

 

Wool-Black 29 42 33 29 33 6
Non-Native Fibers 5 3 0 3 3 2
Wool Tan Twill 29 22 16 22 22 5
Non-Native Fibers 1 5 5 1 3 2

 

001
Native Fibers

001 shirt
Native Fibers

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wool-peach crepe 1 13 4 8 7 5

Non-Native Fibers 3 4 12 6 6 4

Cotton (pink snowflake) 22 35 14 23 24 9
Non-Native Fibers 12 9 14 15 13 3

Cotton Bowling shirt 25 4O 36 38 35 7

Non-Native Fibers 0 0 0 O 0 0
Cotton sage 9m wlyellow 23 27 28 30 27 3
Non-Native Fibers 2 4 4 5 4 1

Cotton-teal knit 66 62 64 27 55 19
Non-Native Fibers 28 69 37 21 39 21
Cotton-maroon knit 343 264 328 444 345 75
Non-Native Fibers 19 6 31 3 15 13
Cotton-white knit 110 112 131 109] 116 10
Non-Native Fibers 11 9 12 151 12 3
Cotton-flannel 26 30 42 36 34 7
Non-Native Fibers 1O 4 6 7 7 3
Denim #1 452 428 402 491 443 38
Non-Native Fibers 5 4 6 4 5 1

Denim #2 21 17 15 23 19 4
Non-Native Fibers 21 17 15 23 19 4
Denim #3 115 114 57 80 92 28
Non-Native Fibers 7 7 13 10 9 3
[Nylon-mesh 30 21 25 2O 24 5
[Non-Native Fibers 8 8 4 10 7 3
[Nylon-greenlgrey 12 11 16 13 13 2
[Non-Native Fibers 7 4 3 7 5 2
[Nylon-creme 7 18F 71 12] 11 4

 

 

 

 

 

 

 

 

RMSI

 

 

Comparison of Raw Mean Shedding Index between the same fabric types

 

Appendix IV

Nylon Fabrics

 

1 2

Fabrics Tested

 

49

 

7!

 

 

Appendix IV (cont’d).

 

 

 

 

Cotton Fabrics

500 4+ .+ e- -221-
400 a
300 '
200 l
100 -

RMSI

 

1 2 3 4 5 6 7 8 9 10
Fabrics Tested

Wool Fabrics

 

RMSI

 

 

 

1 2 3 4 5 6
Fabrics Tested

50

 

A_._4

 

 

 

 

Appendix IV (cont’d).

 

Rayon Fabrics

60

 

50 -
40 ~
30 -.
20 a
10 -

RMSI

 

       

1 2
Fabrics Tested

 

 

51

 

 

 

Appendix V

Categorization of Nineteen Fabrics Tested

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Medium-Low Low
High Shedders High-Medium Shedders Medium Shedders Shedders Shedders
*X>80 80>*X>46 45>*X>26 25>*X>16 *X>15
mfimfl’l/W 710201!-
DE/Wl/I/ WED/MM V/(IV/T WOOL-PLAID W CREME
mmm
RSV/”rt? MVWMSYPAWHV WWW/101%? M1 UNI/55W 677505
WNW/0W7: P01}! MW
mm cvrrmmwr-mz DEN/III! . Bl (IE/UV”
7001K fill/E
WWW/71W”? ACHfl/c m-mnmz SILK-(M
PU! K 516W
SUfDEAW/f

 

 

 

 

 

 

 

 

*X: Denotes Raw Median Shedding Index

Note: For clarification of specific fabrics noted above see Appendix II

52

Appendix VI

Results of Hug and Handshake Transfers

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Number of fibers transferrred: HUG 1 HUG 2 HUG 3 HANDSHAKE
Acrylic 170 165 83 NR
Non-Native Fibers 30 18 12
Poly.Black Suede-knit 757 576 456 3
Non-Native Fibers 113 101 94 5
Poly. Blue Silk-like feel 19 17 16 NR
Non-Native Fibers 87 95 85
Poly. Hot Pink knit 51 47 43 NR
Non-Native Fibers 50 57 53
Poly. Navy blue knit 51 42 37 NR
Non-Native Fibers 60 40 61
Rayon-Daisy Pattern 82 72 70 3
Non-Native Fibers 37 69 65 17
Wool-Plaid 1 00 52 40 NR
Non-Native Fibers 54 30 19
Wool-peach crepe 297 140 1 10 NR
Non-Native Fibers N/A N/A N/A
Cotton (pink snowflake) 44 26 18 NR
Non-Native Fibers 115 80 55
Cotton-teal knit 66 65 58 1
Non-Native Fibers 75 60 40 17
Cotton-maroon knit 542 535 437 2
Non-Native Fibers 54 106 70 15
Cotton-white knit 453 386 191 NR
Non-Native Fibers N/A N/A N/A
Cotton-flannel 135 82 78 N R
Non-Native Fibers 62 28 31
Denim #1 451 372 302 4
Non-Native Fibers 40 37 25 6
Denim #2 81 51 48 NR
Non-Native Fibers 91 55 69
Denim #3 144 89 76 NR
Non-Native Fibers 28 14 20

 

 

 

Appendix VI (cont’d).

 

 

 

 

 

 

 

 

 

 

 

 

 

Wool Tan Twill 22 21 8 2
Non-Native Fibers 65 46 75 13
Nylon-mesh 25 20 1 7 0
Non-Native Fibers 95 115 150 12
Nylon-creme 16 15 12 NR
Non-Native Fibers 50 44 47

 

NR-no result obtained

54

 

 

Appendix VII

Calculation of Mean Transfer

 

DONOR FABRIC MEAN TRANSFER
Acrylic 139
Poly. Black Suede-knit 596
Poly. Blue Silk-like feel 17
Poly. Hot Pink knit 47
Poly. Navy Blue knit 43
Rayon-Daisy Pattern 75
Wool-Plaid 64
Wool Tan Twill l7
Wool-Peach Crepe 182
Cotton (pink snowflake) 29
Cotton-Teal knit 63
Cotton—Maroon knit 505
Cotton-White knit 343
Cotton-Flannel 98
Denim #l 375
Denim #2 60
Denim #3 103
Nylon-mesh 21
Nylon-créme 14

55

 

Appendix VIII
Results of MIN ITAB Statistical Program

Pearson correlation of Hug and RMSI = 0.753

 

 

Fabric Hug RMSI Hug RMSI
Rank Order Rank Order
Acrylic 139 64 14.0 14.0
Poly. Black Suede 596 269 19.0 17.0
Poly. Silk-like feel 17 6 2.5 1.0
Poly.Hot Pink knit 47 57 7.0 13.0
Poly. Navy Blue knit 43 15 6.0 4.0
Rayon—Daisy Pattern 75 56 l 1.0 12.0 a
Wool-Plaid 64 35 10.0 10.0
Wool Tan Twill 17 22 2.5 6.0
Wool-Peach Crepe 182 7 15.0 2.0
Cotton (pink snowflake) 29 24 5.0 7.5
Cotton-Teal knit 63 55 9.0 l 1.0
Cotton-Maroon knit 505 345 18.0 18.0
Cotton—White knit 343 l 16 16.0 16.0
Cotton-Flannel 98 34 12.0 9.0
Denim #1 375 443 17.0 19.0
Denim #2 60 19 8.0 5.0
Denim #3 103 92 13.0 15.0
Nylon-mesh 21 24 4.0 7.5
Nylon-creme 14 l l 1.0 3.0

56

10.

ll.

12.

REFERENCES

. Colyer, A. M. Changes in fiber shedding characteristics as a garment ages. M.A.

Thesis, Michigan State University. 1997.

Coxen, A., Grieve, M. & Dunlop, J. A method of assessing the fibre shedding
potential of fabrics. Journal of the Forensic Science Society. 1992; 32: 151-158.

Grieve, M. Fibres and their examination in forensic science. Forensic Science
Progress. 1990; 4:44-1 I9.

. Greive, M., Dunlop, J. & Haddock, P. Transfer experiments with acrylic fibers.

Forensic Science lntemational. 1989; 40:267-271.
Hatch, K. Textile Science. Minneapolis: West Publishing Company, 1993.

Houck, M. M. Statistics and trace evidence: The tyranny of numbers. Forensic
Science Communications. 1999; 1(3).

Kidd, C. & Robertson, J. The transfer of textile fibres during simulated contacts.
Journal of the Forensic Science Society. 1982; 22:301-308.

Lowrie, C. N. & Jackson, G. Recovery of transferred fibers. Forensic Science
International. 1991; 50:1 1 1.

Mendenhall, W., Beaver, R. & Beaver, B. Introduction to probability and
statistics-10th edition. California: Brooks/Cole, 1999:685-693.

Nickolls, L. The scientific investigation of crime. London: Butterworth, 1956:
39.

Palmer, R. The retention and recovery of transferred fibers following the washing
of recipient clothing. Journal of Forensic Sciences. l998;43(3):502-504.

Parybyk, A. & Lokan, R. A study of the numerical distribution of fibres
transferred from blended fabrics. Forensic Science Society. 1986; 26:61—68.

. Pounds, C. & Smalldon, K. The transfer of fibres between clothing materials

during simulated contacts and their persistence during wear-Part 1. Fiber
transference. Journal of Forensic Science Society. 1975; 15: 17-27.

. Pounds, C. & Smalldon, K. The transfer of fibres between clothing materials

during simulated contacts and their persistence during wear-Part II. Fiber
persistence. Journal of Forensic Science Society. 1975; 15:29-37.

57

15.

16.

17.

18.

Pounds, C. & Smalldon, K. The transfer of fibres between clothing materials
during simulated contacts and their persistence during wear—Part III. A
preliminary investigation of the mechanisms involved. Journal of Forensic
Science Society. 1975; 15:197-207.

Robertson, J ., Kidd, C. & Parkinson, H. The persistence of textile fibres
transferred during simulated contacts. Journal of the Forensic Science Society.

1982; 22:353-360.

Salter, M., Cook, R. & Jackson, A. Differential shedding from blended fabrics.
Forensic Science International. 1987; 33: 155-164.

Siegel, J.A. Evidential value of textile fiber-Transfer and persistence of fibers.
Forensic Science Review. 1997; 9:81-95.

58

llililiiiiliiiiiigiii