Evaluate the claim "we can predict who will become a criminal"

A timeline toolkit for cold case investigations

David Keatley and David D. Clarke

Abstract

Purpose – The purpose of this paper is to outline a variety of related methods for helping with criminal (cold) case investigations. Despite the best efforts of police investigations, many cases around the world

run out of leads and go cold. While many police departments around the world have developed specialist

groups and task forces, academics have also been developing new methods that can assist with

investigations.

Design/methodology/approach – Cold cases, by their very nature, typically comprise incomplete data sets that many traditional statistical methods are not suited to. Groups of researchers have therefore

developed temporal, dynamic analysis methods to offer new insights into criminal investigations. These

methods are combined into a timeline toolkit and are outlined in the current paper.

Findings – Methods from the timeline toolkit have already been successfully applied to many cold cases, turning them back into current cases. In this paper, two real-world cold cases are analysed with methods

from the timeline toolkit to provide examples of how these methods can be applied in further cold cases.

Originality/value – Methods from the timeline toolkit provide a novel approach to investigating current and cold cases. This review provides academics and practitioners with a guide to begin using and

developing these methods and forming successful collaborations with police departments and cold case

task forces. The methods are also suitable for wider groups and to use in their investigations.

Keywords Methods, Cold case, Investigations, Temporal analysis, Sequence analysis, Timeline toolkit

Paper type Case study

Every crime occurs across varying periods of time. One task facing investigators is to

determine the chain of events that occurred during the commission of a crime. Sometimes

this is relatively simple, based on forensic evidence or eyewitness testimony, other times it

may come down to judgement and heuristics. Despite the best efforts of investigators,

criminal cases may run out of leads or evidence to test, and go “cold”. Such cold cases

have gained widespread media interest in recent years, leading to the development of

specialised Cold Case Task Forces, which typically involve groups of specialists from

police detectives to forensic odontologists. These task forces review cases and consider

what new methods or technology may be used to assist with solving or closing a cold case.

Alongside the advancements made in forensic sciences, academics in the social sciences,

such as psychology and criminology, have also developed new methods and processes

that can assist with police investigations and cold cases. However, from the outset it should

be made clear: these methods do not solve cases. They are provided as a means for

academics to understand how temporal methods can be applied to cold cases, as well as

for detectives to approach their cases from a new angle. The timeline toolkit does not

provide a “magic bullet” to solving cases, but it can be used to bridge the gap between

academics in criminology and psychology, and detectives.

Traditional approaches to research in disciplines such as psychology have focussed on

individual risk factors that are “tested” in statistical approaches such as (hierarchical)

David Keatley is based at

Researchers in Behaviour

Sequence Analysis

(ReBSA), Nottingham, UK

and School of Law,

Murdoch University, Perth,

Australia.

David D. Clarke is based at

the Researchers in

Behaviour Sequence

Analysis (ReBSA) and

School of Psychology,

University of Nottingham,

Nottingham, UK.

Received 16 September 2019 Revised 8 November 2019 Accepted 13 November 2019

DOI 10.1108/JCP-09-2019-0039 VOL. 10 NO. 2 2020, pp. 47-63, © Emerald Publishing Limited, ISSN 2009-3829j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 47

regression models or multilevel modelling, for example, in forensic psychology, research

into attitudes and perceptions are commonly conducted. While this provides some

interesting information on the cognitions of students (used as a proxy for offenders), it has

limited applied value. Furthermore, the statistics underpinning such studies (e.g. regression

analyses) posit a number of predictor variables on a single outcome behaviour[1].

Therefore, groups of researchers have turned to novel approaches to understand criminal

behaviours (Keatley, 2018; Taylor et al., 2008) as well as investigating cold cases or crime

scene behaviours (Ferguson, 2015, 2019). The focus of this paper is to outline a series of

methods that academics have been using in the area of forensics and criminology,

collectively termed the Timeline Toolkit, as they share a commonality in focussing on

temporal episodes of behaviours. Each method will be listed and explained in this paper.

Finally, two real-world cold cases will be investigated through two of the “tools” in the

timeline toolkit, crime script analysis (CSA; Cornish, 1994; Leclerc and Wortley, 2013) and

Behaviour Sequence Analysis (BSA; Clarke and Crossland, 1985; Ivanouw, 2007; Keatley,

2018), to show investigators how such approaches may assist with their investigations. The

aim is to bridge the gap between cold case detectives and academics.

Data collected in most controlled, laboratory studies in academia are typically complete

and relatively well-structured. While the researcher may not know the outcome of the study,

they likely know what the data will look like, and missing data are purposefully kept to a

minimum. This makes for clearer data from which results can be seen and conclusions

drawn; however, such methods are less useful in real-world contexts and cases, wherein

the data or evidence are typically unstructured, competing, missing and/or chaotic. Many

cold cases have significant gaps in knowledge or evidence, and attempting to run

conventional statistical techniques would not only be unwise, but in many cases would not

even be possible. While many detectives are open to any additional input and advice

available for their cold cases, they are under intense pressure to solve their cases in a way

that is defensible in court. Many cases that have gone cold lack the forensic evidence

necessary to solve them. It is at this stage, in particular, that methods from the timeline

toolkit can be of most use: providing fresh insight or novel understanding. It should

be made clear, however, that while the methods have proven useful in a number of cold

cases already, they are not a magical remedy to fix or solve all cases. Sound detective work

and expertise is, and will always be, central to solving crimes. The timeline toolkit, however,

offers several methods that can help investigations along the way.

Timeline Toolkit

Everything we do happens over the course of episodes in time. These episodes may span

from a few seconds to many years, but the process for analysing them remains similar.

Timelines are like maps, in the same way that a map can be zoomed-in on for street-level

detail, or zoomed-out of to global scales, maps and directions still follow the same

principles. Within temporal analyses, there are a number of techniques for analysing the

progression of behaviours and events across time. For investigative purposes, this allows

academics to analyse either life histories of potential suspects or victims (Keatley, Golightly,

Shephard, Yaksic and Reid, 2018), or crime-scene-specific behaviours (Fossi et al., 2005;

Keatley et al., 2017; Lawrence et al., 2010). The scale may change, but the directions or

methods remain the same.

Recently, a group of academics have formed an international network of specialists in

methods for temporal analyses: Researchers in Behaviour Sequence Analysis (ReBSA;

www.ReBSA.co.uk). The “timeline toolkit” refers to a number of methods for temporal

analyses that members of ReBSA have either developed or used. Across most of the

methods used in the timeline toolkit, the underlying principle is to analyse the progression of

behaviours or events. To do this, an episode first needs to be broken down, or parsed, into

discrete parts. These parts may refer to behaviours (e.g. stabs victim, moves body), or

PAGE 48j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

events (e.g. graduates school, loses job and divorce), or “chunks of time” in which a count

of how many behaviours occur is measured (e.g. number of burglaries between 10 p.m.

and midnight; number of attacks reported from 2012 to 2013). The parsing process is

typically straightforward; however, it is sometimes not necessarily clear when a behaviour

starts and ends. For example, “shoots a gun” is a relatively discrete behaviour that is clearly

marked in time as beginning (pulling the trigger) and ending (gun fires); however, cognitive

states (e.g. depression) or diffuse, prolonged behaviours (e.g. physical abuse) may occur

over an extended period with start-and-end points less clearly defined. Researchers,

however, have developed the tools to account for and analyse all of these types of

behaviour. The focus of researchers working with timeline toolkit methods is to make the

methods fit the data, not vice versa. Therefore, an aim of the present research is to bridge

the gap between cold case detective work and academic methodology (see Table I, for

examples). Many of the methods in the timeline toolkit are newly developed (e.g. behaviour

fingerprinting; behaviour tracking) and clearly need to be re-tested over multiple cases – as

is made clear by the authors of those methods. However, the underpinning metrics in those

papers is based on well-researched approaches (e.g. Markov modelling). Again, these

methods are not like forensic tests – hence they are not “solving” a case, but the methods

do allow for novel ways of understanding or investigating a crime – perhaps opening a new

line of enquiry for police to explore. Of course, academics should continue developing

these methods and exploring their reliability across different cases[2].

Once behaviours have been parsed or split into separate parts, the next stage is to

categorise or code them. This allows a “dictionary” of sorts to be built so that other

investigators and researchers can immediately see what each behaviour entailed, with a

description, as well as allowing future cases to be added to the database or compared to

existing, known cases. The categorisation stage does typically take longer than the parsing

stage. It is important that behaviours are categorised appropriately so that they are

meaningful in a legal context, as well as for research purposes. For example, if a crime is

classified as “burglary” it may not be immediately clear in legal terms what specific

behaviours have been performed. In the UK, for example, burglary is defined as several

different types or categories, each with specific behaviours. In the USA, “breaking and

entering” defines a similar type of crime to burglary, but with some key differences in the

individual behaviours being performed (Keatley, 2018). This is also why the creation of a

code book is particularly useful, so that future researchers, analysts or officers know exactly

what is meant by each category. There has been a trend in research to use sophisticated

terms to explain concepts, which can sometimes alienate end-users, who either do not have

the time to understand what is meant by the new term, or who may feel underwhelmed when

they realise the sophisticated term relates to something they already knew. Simple category

terms are therefore best in timeline analyses.

Once a criminal episode has been parsed and categorised, there are a number of analyses,

depending on which method has been used. The main underlying principle of all analyses is:

what is the order of behaviours and events across time. An overarching benefit of most of the

methods in the timeline toolkit is that they have been developed with end-users in mind and

outputs are typically shown in a way that non-expert audiences can understand and interpret

results. While the methods are all underpinned by scientific principles and in some cases

advanced statistical analyses (e.g. T-pattern analysis; Magnusson, 2000), the output is simplified

for wider audiences. Alongside the intuitive appeal of a method that takes the experience and

expertise of the detectives involved and builds on it, the presentation of complex results in clear

outputs is a benefit of these methods.

Criminology and the timeline toolkit

All of the methods in the timeline toolkit can be applied to any area of human

behaviour; however, there has been a recent surge in the temporal analyses of crime

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 49

T a b le

I T im

e lin e to o lk it m e th o d s, d e sc ri p tio

n s, a n d a ss o ci a te d a u th o rs

N a m e o f m e th o d

D e s c ri p ti o n

A p p lic a ti o n /E x a m p le

R e fe re n c e s /A u th o rs

a s s o c ia te d

B e h a v io u r s e q u e n c e

a n a ly s is (B S A )

T h is is o n e o f th e m o s t p re v a le n t m e th o d s o f s e q u e n c e

a n a ly s is . It is a v e ry

d a ta -d ri v e n , b o tt o m -u p a p p ro a c h

in w h ic h in d iv id u a l, d is c re te

b e h a v io u rs

a re

p a rs e d ,

c a te g o ri s e d , a n d a n a ly s e d . It is th e fo u n d a ti o n o n

w h ic h o th e r te c h n iq u e s a re

b a s e d /i n te g ra te d (i .e . c a rd

s o rt in g , b e h a v io u r tr a c k in g , b e h a v io u ra lf in g e rp ri n ti n g ,

h id d e n M a rk o v m o d e lli n g )

U s e d to

o u tl in e th e p o s s ib le p ro g re s s io n o f e v e n ts ,

c o n n e c ti n g e v id e n c e s e e n a t c ri m e s c e n e . C a n a ls o b e

u s e d to

a n a ly s e v e rb a lu tt e ra n c e s (e .g . 9 1 1 c a lls , th re a ts

a n d d e m a n d s )

C la rk e a n d C ro s s la n d (1 9 8 5 ), K e a tl e y

(2 0 1 8 ), K e a tl e y , M a ro n o a n d C la rk e

(2 0 1 8 ), M a ro n o e t a l. (2 0 1 7 ), (R e B S A .c o .

u k )

C a rd

s o rt in g

A s e ri e s o f c a rd s a re

u s e d to

d is p la y th e te m p o ra l

p ro g re s s io n o f e v e n ts a n d b e h a v io u rs . T h is h a s b e e n

u s e d in s e lf -h a rm

re s e a rc h p re v io u s ly , th o u g h th e

in te rv ie w in g o f w it n e s s e s c o u ld

b e n e fi t fr o m

th is

te c h n iq u e , w h ic h c a n im

p ro v e re c a ll a n d a c t a s

p ro m p ts fo r fu rt h e r q u e s ti o n in g

U s e fu lf o r w it n e s s /r e la ti v e s w h o m a y fi n d it h a rd e r to

s p e a k a b o u t th e e v e n ts . A n o v e lw

a y o f e lic it in g

in fo rm

a ti o n

T o w n s e n d e t a l. (2 0 1 6 ), W a d m a n e t a l.

(2 0 1 7 )

O rd e rs

o f

a p p ro x im

a ti o n

T h e s im

p le s t s e q u e n c e is a “c h a in re a c ti o n ” – e a c h

e v e n t d e p e n d s o n ly o n th e o n e b e fo re . B u t th in g s c a n

b e m o re

c o m p lic a te d . S o m e ti m e s e a c h e v e n t d e p e n d s

o n th e p a rt ic u la r c o m b in a ti o n o f th e p re v io u s tw o

p re v io u s e v e n ts (o r th e p re v io u s th re e , a n d s o o n ).

T h e s e d if fe re n t p a tt e rn s a re

c a lle d “f ir s t- o rd e r” ,

“s e c o n d -o rd e r” a n d “t h ir d -o rd e r” s e q u e n c e s ,

re s p e c ti v e ly . A n d th e lis t c a n g o o n . T h is p ro p e rt y is

c a lle d th e o rd e r o f th e s e q u e n c e , w h ic h is a c o n fu s in g

te rm

, a s “o rd e r” s o u n d s lik e a n o th e r w o rd

fo r

“s e q u e n c e ”. It m a y h e lp

to th in k o f th e “o rd e r” o f a

s e q u e n c e a s m e a n in g it s “d e g re e o f c o m p le x it y ”

A s ta ti s ti c a ld

e v e lo p m e n t o f s ta n d a rd

B S A in w h ic h

lo n g e r p re d ic ti v e c h a in s c a n b e m a d e . H a s s im

ila r

a p p lic a ti o n a s B S A ; b u t, w it h lo n g e r te m p o ra lc h a in s .

W h ile

th is s e e m s a g o o d a p p ro a c h , it h a s lim

it a ti o n s o f

o v e rf it ti n g o f d a ta

a n d h e n c e w h y la g -o n e , s ta n d a rd

B S A

is p re fe rr e d

M ill e r a n d S e lf ri d g e (1 9 5 0 )

S u rv iv a la n a ly s is

T h is m e th o d is u s e d to

a n a ly s e th e e x p e c te d d u ra ti o n

o f ti m e u n ti la n e v e n t h a p p e n s . E s s e n ti a lly , fr o m

a

g iv e n p o in t in ti m e , h o w lo n g b e fo re

w e s e e a n e v e n t

o c c u r. E x a m p le : a ft e r a h o m ic id e , h o w lo n g u n ti l th e

(s e ri a l) k ill e r m u rd e rs

a n o th e r p e rs o n ? T h is m e th o d

s h o w s th e ra te

w h ic h e v e n ts o c c u r. It c a n b e u s e d to

s h o w d if fe re n c e s b e tw e e n g ro u p s , fo r e x a m p le , d o

s e x u a lly

m o ti v a te d b u rg la rs

re -o ff e n d m o re

q u ic k ly

th a n fi n a n c ia lly

m o ti v a te d b u rg la rs . O r d o y o u n g e r-

a g e d s e x u a lo ff e n d e rs

re -o ff e n d m o re

q u ic k ly th a n

o ld e r- a g e s o ff e n d e rs

a ft e r re le a s e fr o m

p ri s o n

U s e fu lf o r ri s k a s s e s s m e n t (i .e . h o w lo n g a ft e r a th re a t o r

p ri o r a tt a c k b e fo re

th e n e x t a tt a c k ta k e s p la c e )

B ri ff a e t a l. (2 0 1 3 )

P ro x im

it y c o e ff ic ie n ts

A fo rm

o f s e q u e n c e a n a ly s is in w h ic h p ro x im

it y o f tw o

b e h a v io u rs

o r e v e n ts is m e a s u re d . P a rt ic u la rl y u s e fu l

fo r in te rr o g a ti o n o r n e g o ti a ti o n s it u a ti o n s , in w h ic h th e

d e la y o r d is ta n c e b e tw e e n o n e b e h a v io u r a n d th e n e x t

is im

p o rt a n t to

k n o w

U s e fu lf o r in te rv ie w /i n te rr o g a ti o n s to

m a tc h v e rb a ls ty le .

A ls o u s e fu li n lin g u is ti c a n a ly s e s fo r ti m e b e tw e e n e v e n ts

C o rn e r a n d G ill (2 0 1 9 ), P .J . T a y lo r (2 0 0 6 )

(c o n ti n u e d )

PAGE 50j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

T a b le

I

N a m e o f m e th o d

D e s c ri p ti o n

A p p lic a ti o n /E x a m p le

R e fe re n c e s /A u th o rs

a s s o c ia te d

B e h a v io u r tr a c k in g

T h is is th e fi rs t m e th o d to

c o m b in e te m p o ra la n a ly s e s

(B S A ) w it h g e o g ra p h ic p ro fi lin g in a n a tt e m p t to

s h o w

th e lik e lih o o d o f b e h a v io u rs

in s p a c e

U s e d in c o ld

c a s e s to

m a p p o s s ib le ro u te s ta k e n b y

v ic ti m

a n d p e rp e tr a to r. E s p e c ia lly

u s e fu li n c o n ju n c ti o n

w it h fo re c a s ti n g m e th o d s

K e a tl e y e t a l. (2 0 1 9 )

B e h a v io u ra l

fi n g e rp ri n ti n g

T h is a p p ro a c h b u ild

s o n th e C ri m e L in k a g e re s e a rc h

b y p ro v id in g a w a y o f m e a s u ri n g th e s im

ila ri ty b e tw e e n

tw o b e h a v io u ra lc h a in s . F o r e x a m p le , if 3 w o m e n a re

fo u n d m u rd e re d , in v e s ti g a to rs

m a y w o n d e r w h e th e r

th e y a re

lo o k in g fo r 1 , 2 , o r 3 k ill e rs . If w e lo o k a t th e

s e q u e n c e o f b e h a v io u rs

in e a c h c a s e , w e c a n th e n s e e

w h e th e r th e b e h a v io u ra lf in g e rp ri n ts o f e a c h c ri m e

s c e n e m a tc h

N o t o n ly fo r lin k in g s u s p e c te d s e ri a lk ill e rs , b u t c a n a ls o

b e u s e d to

“f in d ” s im

ila r c ri m e s a llo w in g in v e s ti g a ti o n s to

lo o k a t th o s e c ri m e s a n d c o m p a re

M O , a n d s o lv a b ili ty

fa c to rs .

T h is m e th o d a ls o a llo w s ti m e lin e s d e v e lo p e d v ia

in te rr o g a ti o n to

b e c o m p a re d to

k n o w n fa c ts to

h ig h lig

h t

d if fe re n c e s

K e a tl e y a n d C la rk e (2 0 1 8 )

T im

e s e ri e s a n a ly s is

T h is a p p ro a c h ta k e s c o n ti n u o u s v a ri a b le s (l ik e a g ra p h

o f c ri m e ra te s ) a n d ta k e s th e s e ri e s a p a rt th e w a v e s

in to

c y c le s . It p u lls

o u t tr e n d s in th e c o n ti n u o u s ly

fl u c tu a ti n g v a ri a b le s

M o re

u s e fu lf o r c o m p a ri s o n o f la rg e r d a ta

s e ts o f lo n g e r

ti m e lin e s (i .e . c o m p a ri n g h o w m a n y m u rd e re rs

h a d ri s k

fa c to r “x ” in th e ir c h ild

h o o d )

B a k e m a n a n d G o tt m a n (1 9 8 6 )

H id d e n M a rk o v

m o d e lli n g

In B S A , W h ile

w e m a y o b s e rv e B e h a v io u r A (t a lk in g

c a lm

ly ) a n d B e h a v io u r B (s m ili n g ), th e y a c tu a lly

re fl e c t

a n u n d e rl y in g s ta te

(S ta te

1 : p e rp e tr a to r c a lm

), a n d

B e h a v io u r C (s h o u ti n g ) m a y re fl e c t a d if fe re n t

u n d e rl y in g s ta te

(S ta te

2 : p e rp e tr a to r a n g ry ). W h e re a s

A a n d B m a y c h a n g e (i .e . th e o b s e rv e d b e h a v io u rs

m a y b e d if fe re n t, s u c h a s D – la u g h in g ) th e u n d e rl y in g

s ta te

m a y re m a in th e s a m e . S im

ila rl y , B e h a v io u r D

(p u n c h e s ) m a y b e o b s e rv e d fo llo w in g A a n d B ; b u t, D

re m a in s a m a n if e s ta ti o n o f s ta te

2 (a n g ry )

U s e fu lf o r u n d e rs ta n d in g u n d e rl y in g s ta te s ; h o w e v e r,

n o t c u rr e n tl y u s e d in c o ld

c a s e in v e s ti g a ti o n s . M o re

d a ta

a re

re q u ir e d a n d th e “h id d e n ” a s p e c t is ty p ic a lly

o ff -

p u tt in g to

in v e s ti g a to rs

E d d y (2 0 0 4 ), Z u c c h in ie t a l. (2 0 1 7 )

C lu s te re d B S A

A n a ly s is o f h ig h e r- o rd e r s e q u e n c e s s e e m s to

re q u ir e

c o lo s s a ln u m b e rs

o f d a ta . F o r in s ta n c e , a s e q u e n c e

w it h 2 5 k in d s o f e v e n ts w o u ld

n e e d a ta b le w it h 6 2 5

c e lls

fo r a fi rs t- o rd e r a n a ly s is , b u t 1 5 ,6 2 5 c e lls

fo r

s e c o n d o rd e r, a n d 3 9 0 ,6 2 5 c e lls

fo r th ir d o rd e r.

C le a rl y , a “c o m b in a to ri a le x p lo s io n ” h a s s e t in .

D a w k in s (1 9 7 6 ) s o lv e d th is p ro b le m

w it h a c le v e r

m e th o d th a t fi n d s s h o rt c h a in s fi rs t, a n d th e n h o m e s in

o n th o s e to

s e e if th e y a re

th e k e rn e lo f lo n g e r c h a in s –

a b it lik e g ro w in g a c ry s ta lf ro m

a s m a ll in it ia l “s e e d ”.

B e c a u s e th e s e a rc h fo r p a tt e rn s is s o s e le c ti v e , th e

re q u ir e m e n t fo r d a ta

o n ly g ro w s g ra d u a lly

a s lo n g e r

a n d lo n g e r c h a in s a re

c o n s id e re d , a n d th e d e te c ti o n o f

lo n g -c h a in p a tt e rn s b e c o m e s a p ra c ti c a lp

o s s ib ili ty

a g a in

A s ta ti s ti c a ld

e v e lo p m e n t o f th e B S A a p p ro a c h to

s h o w

h o w b e h a v io u rs

c lu s te r to g e th e r in to

la rg e r s e q u e n c e s

D a w k in s (1 9 7 6 )

(c o n ti n u e d )

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 51

T a b le

I

N a m e o f m e th o d

D e s c ri p ti o n

A p p lic a ti o n /E x a m p le

R e fe re n c e s /A u th o rs

a s s o c ia te d

C ri m e s c ri p t a n a ly s is

In c o n tr a s t to

B S A , th is is a to p -d o w n a p p ro a c h b a s e d

o n p s y c h o lo g ic a l s c ri p t th e o ri e s . T h is a p p ro a c h is

m o re

q u a lit a ti v e a n d o ft e n e a s ie r fo r p ra c ti ti o n e rs

to

u n d e rs ta n d a s it d o e s n o t n e e d s ta ti s ti c a lt e c h n iq u e s .

It a ls o w o rk s p a rt ic u la rl y w e ll c o m p a re d to

B S A o n d a ta

o r c a s e s w h e re

n o t a ll in fo rm

a ti o n is k n o w n . It w o rk s

b e s t w h e n o th e r c a s e s c a n b e u s e d to

b u ild

a “s c ri p t”

a n d th e n th e c a s e b e in g in v e s ti g a te d c a n b e

c o m p a re d to

th o s e s c ri p ts /t h o s e s c ri p ts c a n b e u s e d

u n d e rs ta n d th e c o ld

c a s e

A s o u tl in e d in th is p a p e r, u s e fu lf o r c a s e s in w h ic h

d e ta ile d s p e c if ic in fo rm

a ti o n is n o t a v a ila b le . It p ro v id e s

a n in it ia lt e m p o ra la p p ro a c h to

c ri m e in v e s ti g a ti o n s . It

a ls o in v o lv e s p o lic e in th e d e v e lo p m e n t o f ti m e lin e

c h a in s , w h ic h w o rk s a s a u s e fu ls ta rt in g p o in t. A llo w s

re s e a rc h e rs

to b u ild

d a ta b a s e s o f C S A c h a in s

B e a u re g a rd

e t a l. (2 0 0 7 ), E k b lo m

a n d

G ill (2 0 1 6 ), H o m e le t a l. (2 0 1 4 ), K e a tl e y

e t a l. (2 0 1 9 ), B . L e c le rc

a n d W o rt le y

(2 0 1 4 ), B e n o it L e c le rc

e t a l. (2 0 1 1 )

T -p a tt e rn

a n a ly s is

T h is a n a ly s is a llo w s fo r m u lt ip le c o n c u rr e n t a n d

s e q u e n ti a l b e h a v io u rs

to b e a n a ly s e d a c ro s s ti m e . T h e

p a tt e rn s (e .g . X Y Z ) c a n b e a n a ly s e d to

s h o w h o w

fr e q u e n tl y th e y o c c u r a c ro s s ti m e , a n d w h e th e r th e y

fo rm

p a rt o f a la rg e r h ie ra rc h y o f s e q u e n c e s (e .g . (A !

(X Y Z )) ). T -p a tt e rn

a n a ly s is c a n fi n d s e q u e n c e s th a t a re

n o t im

m e d ia te ly a p p a re n t o r o b v io u s , e v e n to

th e

tr a in e d e y e – e s p e c ia lly

in c a s e s w h e re

s e q u e n c e

p a tt e rn s a re

s p re a d o v e r ti m e . T h e a n a ly s is d o e s le a d

to c o m p le x o u tp u ts ; h o w e v e r, tr a in e d e x p e rt s c a n

fa c ili ta te

u n d e rs ta n d in g th ro u g h c le a re r e x p la n a ti o n s

th a n th e c o m p u te ri s e d o u tp u t c u rr e n tl y g iv e s

U s e fu lf o r in te rr o g a ti o n a n a ly s e s . W h ile

th e a c a d e m ic

lit e ra tu re

is ty p ic a lly

u n fa v o u ra b le to

d e c e p ti o n

d e te c ti o n , th e a b ili ty

to c o m b in e v e rb a la n d n o n v e rb a l

c u e s is u s e fu l

M a g n u s s o n (2 0 0 0 ), M a g n u s s o n e t a l.

(2 0 1 6 )

In d ic a to r w a v e s

T h is a p p ro a c h a ls o a llo w s m u lt ip le c o n c u rr e n t

b e h a v io u rs

a n d e v e n ts to

b e m e a s u re d a n d w a v e s a re

p lo tt e d to

in d ic a te

w h e n a fu tu re

b e h a v io u r m a y o c c u r.

F o r e x a m p le , if w e s e e b e h a v io u rs

x , y a n d z o c c u rr in g

a b o v e th e le v e lw

e w o u ld

e x p e c t to

s e e b y c h a n c e ,

d o e s th a t m e a n th a t a te rr o ri s t a tt a c k is im

m in e n t o r s ti ll

m a n y m o n th s a w a y

A m o re

u s e r- fr ie n d ly v e rs io n o f T -P a tt e rn

A n a ly s e s . T h e

o u tc o m e s a re

m u lt ip le w a v e s . M o re

u s e fu lf o r m u lt ip le

c a s e s (e .g . lif e h is to ry

a n a ly s e s )

K e a tl e y a n d C la rk e (2 0 1 8 )

T e m p o ra ln e tw o rk

a n a ly s is

(S o c ia l) n e tw o rk

a n a ly s is o u tl in e s th e re la ti o n s h ip

b e tw e e n in d iv id u a ls , fo r in s ta n c e w h o g a n g s in te ra c t.

F ro m

th e s e n e tw o rk s , h ie ra rc h ie s o f c o n tr o l o r p o w e r

c a n b e s e e n , in fl u e n c e c a n b e n o te d . T h is is e x tr e m e ly

im p o rt a n t fo r in v e s ti g a ti o n s ; h o w e v e r, re c e n t a d d it io n s

h a v e a ls o b e g u n to

m a p h o w th e s e n e tw o rk s g ro w o r

c h a n g e o v e r ti m e . T h e te m p o ra ln e tw o rk

a n a ly s e s c a n

a ls o b e u s e d to

s h o w h o w c o -r e la te d b e h a v io u rs

a t th e

s c e n e o f a c ri m e c h a n g e o v e r ti m e (e .g . if g u n s a n d

g a g s a re

u s e d in e a rl y c ri m e s , d o e s th is c h a n g e to

k n iv e s a n d ta p e la te r o n , o r is th e n e tw o rk

m o re

in tr ic a te ly lin k e d )

T h e b a s is o f th is h a s b e e n u s e d to

s h o w th e re la ti o n s

b e tw e e n p e o p le in c a s e s . C o ld

C a s e F ile s a re

fi lle d w it h

n a m e s (s u s p e c ts , w it n e s s e s , re la ti v e s , ti p -o ff s , e tc .) .

T h e re

a re

ty p ic a lly

m o re

n a m e s th a n d e te c ti v e s c a n

k e e p tr a c k o f, s o p ro v id in g a n a n a ly s is o f th o s e n a m e s

a n d lin k s to

th e c a s e a n d e a c h o th e r p ro v id e s a h e lp fu l

o v e rv ie w .

G ill e t a l. (2 0 1 4 ), K e m p e e t a l. (2 0 0 3 ),

K re b s (2 0 0 2 )

(c o n ti n u e d )

PAGE 52j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

T a b le

I

N a m e o f m e th o d

D e s c ri p ti o n

A p p lic a ti o n /E x a m p le

R e fe re n c e s /A u th o rs

a s s o c ia te d

M a tr ix fo re c a s ti n g -B S A

T h is c o m b in e d a p p ro a c h o f m a tr ix fo re c a s ti n g a n d

B S A h a s b e e n u s e d o n c o ld

c a s e s to

p ro v id e a

s u m m a ry

o v e rv ie w o f th e c a s e p ro g re s s io n . In

p a rt ic u la r, th e m a tr ix fo re c a s ti n g o u tl in e s d e te c ti v e s ’

ju d g e m e n ts a n d in v e s ti g a ti v e a p p ro a c h e s , w h ic h a re

m a d e a b o u t b e h a v io u rs

a n d e v e n ts a c ro s s th e

ti m e lin e , o u tl in e d in B S A . T h e c o m b in a ti o n o f th e s e

m e th o d s a llo w s a s tr u c tu re d ti m e lin e o f in v e s ti g a ti v e

le a d s to

b e q u ic k ly s u m m a ri s e d , a s w e ll a s p o in ts o f

p o te n ti a lf u tu re

in v e s ti g a ti o n to

b e s e e n . T h e s o lv a b ili ty

s c a lin g in th is m e th o d a ls o m a k e s it u s e fu lf o r

p ri o ri ti s in g fu tu re

c a s e s

T h is is o n e o f th e m o s t u s e d a p p ro a c h e s . T y p ic a lly

in

c o ld

c a s e s , s e q u e n c e s o f e v e n ts a re

n o t k n o w n , th is

m e th o d a llo w s a s y s te m a ti c a p p ro a c h to

m a p p in g th e

p ro p o s e d p a th w a y s , re q u ir in g c o ld

c a s e te a m s to

o u tl in e w h y th e y p ro p o s e th e s e q u e n c e , w h a t e v id e n c e

th e re

is to

s u p p o rt it , a n d w h a t te s ts c o u ld

b e d o n e to

s u p p o rt it

K e a tl e y a n d C la rk e (2 0 1 9 )

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 53

(Keatley, Marono and Clarke, 2018; Marono et al., 2017, 2018; Taylor et al., 2008),

including cold case investigations. One aspect of crime that makes it particularly

suitable for timeline toolkit methods is that the commission of a crime occurs in a fairly

clear cut episode. There is a relatively clear start-and-end point that can be outlined,

giving a clear window for analyses[3]. There are methods to help with many cold

cases, from single-offender homicide, to serial killers, through to serious sexual

assault and rape cases. The methods can be applied to cases with relatively few

known facts; however, obviously a more complete timeline can help with analyses.

There are also methods that review and help with the investigative procedure itself,

outlining detectives’ decision-making processes and the progress of the cold case. A

further benefit of a temporal approach to crime analysis is that further statement

analysis of interrogations can be compared to developed timelines, which can help

investigations.

Within timeline analyses, a common concern is that methods focus too narrowly on

behaviours, and the psychology is taken out of the analyses. For example, in BSA (Keatley,

2018), the output is a statistical flow diagram of chains of behaviours and events. While this

is informative and useful, it could be considered too descriptive and lacking in the richness

of detail that criminal psychologists prefer to see. It would be wrong, however, to think that

temporal methods do not apply psychological concepts to the models that are built.

Instead, classic psychological and criminological explanations can be used to explain the

chains of behaviours that are seen in the outputs. For example, in a study on interpersonal

violence in the night-time economy, Taylor et al. (2017) conducted a BSA of the

antecedents of violence. This was, on the surface, a descriptive representation of the typical

chains of behaviours that occur preceding a violent outburst in a bar or club. However, this

is not where the analyses end; instead, once a BSA has been conducted on the data, it is

then the job of psychologists, criminologists and/or investigators to discuss and theorise on

why those chains are most likely to occur – what motivations, intentions or goals were

underlying those behaviours, or what contextual cues or elements lead to them. In some

instances, it is a simple explanation: someone knocking your drink/spilling their drink on you

is seen as a provocation or precipitator (Wortley, 2016) to a violent outburst; in other cases a

more complex set of heuristics or rational decision making is needed to understand the

motivations or rationale of the behaviours.

BSA is one of the most basic methods for building chains of behaviours. BSA has proven

useful in raising awareness of a temporal analysis approach to research, but it has also

come under criticism for reducing complex reasoning to descriptive chains of behaviours

and for being overly simplified (Keatley, 2018). In terms of the first issue, as outlined,

descriptive methods such as BSA should be seen as a framework from which to build

psychologically informed explanations for the patterns or sequences. In terms of the second

issue, over simplification of chains of behaviours into step-by-step chains, there are

methods that allow for more complex chains of behaviour to be shown. Figure 1 provides a

visual display of the issue. On the left is a typical BSA flow diagram (called a state transition

diagram). This shows how one behaviour or event (A) is followed by a second, or sequitur

behaviour (B), and in turn (B) is followed by (C) and so on. In the basic form of BSA, a lag-

one BSA, behaviours are analysed as pairs (A!B, B!C, etc.). Clearly, real-life interactions are seldom this clean-cut and clear. It is entirely possible for me to (A) shout at someone,

while simultaneously, (B) hit someone. This is shown on the right hand side of Figure 1, in

which (A) and (B) co-occur, possibly leading to (C) the person falls down. Owing to some

ground-breaking work by Magnus Magnusson and his team in the Human Behaviour

Laboratory and Research Group on Methodology for the Analysis of Social Interaction, a

method exists that allows for the analyses of concurrent and sequential behaviours: T-

pattern analysis. While the computer outputs of T-pattern analysis are typically much more

complicated to read, and require expert knowledge, the same underlying principles apply

and outputs can be interpreted for wider understanding.

PAGE 54j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

In contrast to timeline analysis tools that reduce the complexity of human rational decision

making to descriptive timelines, there are other methods in the timeline toolkit that analyse

sequences of events from a “top-down” approach. Essentially, rather than playing the role

of “naı̈ve observer” some methods instead build an outline or script of behaviours and

events, based on expertise, awareness and experience. CSA (Beauregard et al., 2007;

Ekblom and Gill, 2016; Leclerc and Wortley, 2014) attempts to explore chains of behaviours

in a similar way to BSA (the two share many similar features, terms and approaches);

however, CSA was originally based on script theory, from cognitive psychology. A script, in

psychological terms, is explained as an individual’s knowledge of how to act and react in

various contexts. The usual example is given by reference to a restaurant. When someone

enters a new restaurant, they typically know how to behave and act based on experience of

being in similar establishments. In the same way that a person knows how to behave in a

new restaurant, because they have a “restaurant script”, criminologists posit that criminals

follow a similar script when planning and commissioning their crime (Keatley et al., 2019).

Whereas BSA focusses on the behaviours observed and from the analyses, psychological

explanations can be given, CSA brings criminal’s motivations and rationality into focus

much earlier on in the analyses. Criminals, for the most part, act out of a rationality or owing

to psychological principles of routines and habits. Understanding their rationale, habits and

intentions informs the development of a script to account for their behaviours.

Application of the timeline toolkit to cold cases

Below are two examples of how the timeline toolkit can be used to assist with cold case

investigations. It should be noted from the outset that neither method, by itself, will solve a

case. That is not their purpose and not their value. The methods, however, can provide

detectives with a new approach to investigating crimes. The methods outlined in the

timeline toolkit have been applied to cold cases in Canada (notably the bombing of

Canadian Pacific Flight 21), and in the USA (notably in Rhode Island Cold Case Task Force

cases[4]). The two approaches and two cases are also chosen to show two different

approaches to sequence analysis. In the first case, a “known sample” approach is used to

collect cases from which to draw likely chains and provide an overview of how the cold case

crime was commissioned. In the second case, BSA is used to show how detectives do not

necessarily need to gather a large database of known cases, instead they can build a

sequence chain from information seen at the crime scene and detectives’ views. These are

similar and connected methods, but take slightly different approaches to sequencing. For

obvious reasons, some parts of the cases have been omitted from this review, so as not to

damage the integrity of the case. The cases are used primarily for illustrative purposes.

To give the methods outline full credibility, real-world case reports and summaries of the

investigative process would be ideal. However, many detectives are understandably

concerned with opening case files to non-law enforcement. The integrity of a case,

Figure 1 Typical BSA output and T-pattern analysis

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 55

including names of suspects and crime scene information, is central to the types of

analyses in the timeline toolkit; however, if current cases were outlined here it may

jeopardise the cases (and would certainly end collaborations between the authors and

detectives)[5]. Therefore, public cases are outlined below and timeline toolkit approaches

re-applied to them as an example of how the methods work. This is meant as an example

for detectives, as well as a tutorial for academics who may know of these methods, but had

not previously thought about them in terms of cold case investigation.

Crime script analysis of a cold case: Wayne Greavette (WG)

On 12 December 1996, WG received a parcel in the mail. Inside were a typewritten letter

and a flashlight. When WG switched the flashlight on, it exploded, instantly killing him. The

letter offered few clues as to the identity of the mail bomber, and despite intensive

investigation the WG case went cold. From the outset, it should be made clear that the

author(s) have not directly worked on the WG case, nor had access to all of the police files.

However, so much of the case has been made public knowledge that it serves as a good

case to openly analyse with CSA.

To apply CSA to the WG case requires a two-pronged approach. First, everything about the

WG case must be analysed to create as coherent and complete a timeline as possible.

This is where access to police files is valuable. In as much detail as is possible, WG’s

background leading up to the event should be examined and chronologically set-out. This

is the typical investigative procedure and such information can be obtained from (cold)

case police files. Alongside this, a number of other, similar cases should also be obtained

and timelines analysed[6]. CSA allows a temporal flow of behaviours and events to be

analysed, rather than a “snapshot” approach that portrays crime as a single event in space

and time (Keatley, 2018). For example, rather than simply label WG as a “mail bomb”, CSA

attempts to place the criminal act within a longer timeline of events and behaviours

surrounding it. Scripts can then begin to be built based on the observed stages or scenes

found in the timelines.

There have been a number of relatively well-known mail bombings over the past few

decades, from which a CSA can be built. Criminals that have sent mail bombs, from Ted

Kaczynski to Guido Amsel, are investigated and their life history and in particular criminal

undertakings are outlined. The focus here is not to find cases that are linked to the current

case – we are not proposing that Kaczynski is WG’s killer; but, similar cases provide a

framework to understand the cold case under investigation. For example, do the majority of

mail bombers hold a personal grudge against the person targeted, or is the letter sent to

“random” addresses? Does the mail bomber know the target in vivo or does the use of mail

require that only the victim’s name and address is known? Is the bomb sophisticated or

relatively simple? How did the criminal learn how to make the bomb? Many detectives

typically feel they have this type of knowledge implicitly stored in their memory, and attempt

to draw from experience or anecdotes when confronted with a new (or cold) case. However,

memory can be biased by various heuristics such as availability – if something comes to

mind quickly, because it is harrowing or horrific, for example, or because it is simply a well-

known case, such as the UNABOM, then this may make us think it is more regularly

occurring than it might be (Roach, 2019). Therefore, a benefit of the CSA approach is to

clearly outline as many similar cases as possible, and from that derive a prototypical script

for the crime under investigation.

To understand more about the WG case, we need to look into the crime scripts of similar

mail bombers. From these cases, we begin to build a crime script for mail bombers (see

Figure 2)[7]. Clearly, there are a number of precipitators for mail bombings, including

financial/business problems, mental issues and personal grievances or grudges towards a

person. These precipitators are the antecedents of the pathway that eventually leads to mail

bombing. Following on from the precipitators is the development of motivations, which can

PAGE 56j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

be broadly split into financial/business and anger/retaliation. WG had a history of “bad”

business ventures and relationships. WG is alleged to have had a history of extramarital

affairs, which may have led to the personal grievance precipitator, and later motivation for

retaliation. Also, the recent purchase of a natural water spring may indicate a

conservationist/anti-industrialist motivation. There is a third motivation in bombings –

suicide; however, in the case of mail bombs the criminal is not delivering and detonating the

bomb in-person so this motivation is not included in this case. The next scene or stage

involves the “planning” of the crime.

There are several key pieces of knowledge that must be gained before a mail bombing can

be completed. Knowledge of the victim’s address must be gained – though this is relatively

straightforward. More importantly in investigative terms: how to build a (mail) bomb must be

researched. Experts in explosives can usually give indication of the complexity of the

device and the likely skill level necessary. Experts involved in the WG case agree generally

agree that the bomb indicated a low-level sophistication, which could be learnt from online

or media sources (books magazines). Within this stage, learning the address of the victim

needs to occur, and police files indicate in the weeks leading up to the bombing a number

of false calls were made to the Greavette household, attempting to attain their address.

There is also an incident involving a truck driving onto their property and being chased

away. The instrumental initiation stage involves the attainment of materials necessary to

make the bomb. The casing and packaging for the bomb need to be obtained – in the WG

case a flashlight was used, and appeared to be new, sent as a gift. Explosives experts can

look at the residue to consider chemical composition of the bombs. Whether the materials

are over-the-counter products combined or specialised compounds can give insight into

the expertise of the bomb maker, as well as investigative leads. It is at this stage a letter or

manifesto may also be written. In the WG case, the letter offers some opportunity to forensic

linguistic analysis, including questions such as: why send a letter at all? Why name certain

people in the letter? The instrumental actualisation stage is the point at which the mail bomb

is sent. Finally, the exit or end stage refers to what happens next. Does the mail bomber

continue to send further bombs (i.e. Kaczynski); do they commit suicide; do they consider

the grudge settled; or do they receive some financial gain from the bombing.

The CSA approach to cold cases provides investigations with a sample of known, solved

cases to compare to. This can provide guidance in a number of ways. First, other methods

used to solve the known crimes can be considered in reference to the cold case. Detectives

who worked on the previous cases can also be consulted to gain an expertise that may be

of assistance. Also, the CSA provides an overview of the crime commission process, which

may provide insights into the rationale and planning of the criminal in the cold case. For

Figure 2 CSA of mail bombing

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 57

example, there is speculation in the WG case that there may be a financial motivation,

owing to the purchase of the water springs that WG and his family had moved on to.

Financial motivations have been seen in similar mail bombing cases. However, personal

grievances are also prevalent in mail bombing crimes. With a script of mail bombings being

built, academics can then compare which of the pathways overlaps most with the WG case.

Of course, criminals do not always function by way of prototypical scripts, and this is why

“matching” in terms of BSA and CSA is not the same as DNA matches. This is also why the

timeline toolkit does not solve cases, per se. But, CSA does provide more investigative

opportunities in terms of understanding typical scripts of similar cases, and new interview

options.

Behaviour sequence analysis of a cold case: Easey Street

On 10 January 1977 in Melbourne Australia, two women were attacked in their home. Susan

Bartlett (SB) and Suzanne Armstrong (SA) were repeatedly stabbed and killed, while SA’s

child remained unharmed in his cot. The bodies of the women and the (still living) child were

not found for three days. Within that time, multiple people confessed to having entered the

house, though none claim to have seen the two women’s bodies or heard any noise from

the unattended child. The case remains unsolved and is referred to as “Easey Street”,

owing to the street on which the house was located. Over the years, numerous suspects

have come into the investigative spotlight, though no arrests have been made. A coronial

inquest in Australia made many of the details of the case publicly available, and personal

communication with reporters and those close to the case have revealed further details[8].

Within the BSA literature, there are a number of publications on sexual assault and rape

cases (Fossi et al., 2005; Lawrence et al., 2010) as well as homicides (Keatley, Golightly,

Shephard, Yaksic and Reid, 2018)[9]. In seminal work conducted by Fossi et al. (2005)

Lawrence et al. (2010), police reports of stranger sexual assault and rape cases in the

bedroom were analysed. Behavioural and verbal sequences were analysed to show the

typical progression or chains that occurred in these cases. Also, Keatley, Marono and

Clarke (2018), Keatley et al. (2019) and Keatley, Golightly, Shephard, Yaksic and Reid

(2018) have conducted research on the life history of serial killers, including sexual

homicides. Keatley et al. showed the life histories of various serial killers, which can be used

to begin to build a “profile” of the type of killer that may be responsible for crimes like the

ones observed in a cold case. This can assist with suspect prioritisation, but should be

done very cautiously to ensure suspects are not “ruled out” carelessly. There is, therefore, a

large database of similar type crimes to those seen in the Easey Street murders. This can

provide cold case investigations with information to fill in what might have happened, which

can then be used to analyse statements made by suspects, to analyse whether their

account is valid. However, large databases are not necessary for timeline toolkit methods,

which can with databases of very few.

There is a huge repertoire of expertise and experiential knowledge in cold case task forces.

It would be unwise, and actually somewhat offensive, to dismiss this expertise entirely.

Therefore, one method for understanding what happened in a cold case such as the Easey

Street murders is to take an account of each detective’s perceptions of what occurred.

Several detectives may believe the intruder was likely to have entered through the front

door, while other detectives may consider the back door or the side window as possible

entry points. Once inside the home, detectives may vary on the sequence of actions

undertaken: who was killed first? When did the attacker go to the bathroom? Did he return to

the victims to move or reposition them (staging the crime scene)? Typically in cold case

reviews these competing views are debated and the lead investigator may attempt to attain

a general consensus; however, BSA can do that in a more systematic approach.

More recently, researchers have linked BSA with other timeline toolkit approaches, such as

Forecasting, to provide an outline of rationales for sequential steps. For instance, if a

PAGE 58j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

detective thinks that a body was moved or repositioned, and the crime scene was staged,

then the reasons for thinking this are noted – which in itself becomes a highlighted issue for

further discussion. Similarly, a combined approach allows investigators to see if any

evidence can be (re)tested with newer methods, such as the MVAC or advancements in

DNA testing. This can provide a quick overview of the cold case for future investigators who

can quickly see a sequence chain showing consensus or divergence in detectives’ views,

as well as areas for further (forensic) investigation.

BSA has also been developed to account for spatio-temporal chains, in a method called

behaviour tracking (Keatley et al., 2019). Behaviour tracking combines the approaches of

BSA with geographic profiling to show the most likely sequence of events through a

building, for example, rather than in a state space, as BSA is normally positioned. As shown

in Figure 3, this can then provide a diagrammatic map of the likely chain of events as a

criminal moves around a home. In the Easey Street case, investigators know certain

behaviours have to occur in certain locations (i.e. washing off blood in the bathroom,

attacking SA in her bedroom and attacking SB in the hallway). By taking the timeline

accounts of detectives, which incorporate the various points of evidence in the property,

pathways can be shown around the home and convergences of opinion clearly outlined.

While this approach is typically better portrayed in an animated presentation, it can still be

reproduced in printed format for police files and future investigators to get an overview of

the logistics and layout of the case in geographic terms.

In Figure 3, the behaviour tracking would show a convergence of opinions that the suspect

entered through the main front door (node “1”) and proceeded directly to SAs bedroom,

wherein he attacked her (node “2”). After this, the suspect went back into the hallway where

he was confronted by SB (node “3”). After attacking SB, the suspect then went to the

bathroom (node “4”) before leaving the premises (node “5”). However, some detectives

may believe the suspect left through the rear door (node “6”). Of course, this is an

oversimplified model of behaviours and movements. There are a number of additional

behaviours that took place in the property, and each “node” is actually representative of a

number of behaviours (such as stab wounds, sexual assault, movement of items, etc.).

However, for illustrative purposes, a diagram can be quickly viewed and embellished with

any considerations of the detectives or consultants. The diagram can easily be annotated

with additional information, such as the existence of blood stains around the bodies and or

on the walls, and annotations regarding inferences that can be drawn from this information.

Key evidence (tested and untested) can also be positioned on the behaviour tracking map,

as a guide for future investigators stating where evidence was collected.

Figure 3 Behaviour tracking analysis of movements around 147 Easey Street

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 59

Conclusions

The main take-home message of this paper is that there are a number of methods that have

been applied to cold cases. The methods provide novel approaches to understanding the

sequence of behaviours that occurred in the crime scene. The majority of methods actively

incorporate detectives’ expertise and foster a parsimonious working relationship between

academics and police practitioners. There are of course shortcomings of the methods in the

timeline toolkit. Notably, when analysing criminal behaviours, detailed accounts are

needed, which are sometimes missing or not possible in cold cases. The methods outlined

cannot be used in all cases, if there is no information available. Missing person cases,

which typically lack much information, are less suited to the current methods. While there is

no “minimum” amount of information required to run the analyses, the less information

known, the vaguer the analyses become. In terms of verbal analyses: if BSA, for example, is

being used in statement analyses of interviews (e.g. Keatley, Marono and Clarke, 2018),

then the longer the interview is, or the more interviews there are, then the more detailed the

analyses become.

The cases outlined in the present paper are indicative of the types of analyses possible

using these methods; however, with unrestricted access to police files, the methods can be

developed and clearer timelines created. Academics wishing to work with detectives are

encouraged to use the methods outlined here and begin building databases for

investigative purposes. Providing detectives with a resource they can use to help

understand or investigate their case, using terms and behaviours they intuitively understand

and can see the overlaps with, is the first step to building bridges. Similarly, these methods

are not “plug and play” and academics in the field are required to run the analyses

alongside detectives. Indeed, a number of groups involving academics and police

practitioners have formed, such as ReBSA, Vidocq, Cold Case Review@ and the American

Investigative Society of Cold Cases. Further academics and police practitioners are

encouraged to join these groups.

Notes

1. Please note, references are omitted here so as not to unfairly critique an individual’s research

when the issues are more widespread in forensic psychology and criminology literature.

2. For example, in missing person cases, many of the methods in the timeline toolkit would have

limited use, as so much of the investigative information is missing. Clearly, the more information

about the crime that is known (in terms of victimology crime scene behaviours) the more useful the

timeline toolkit method becomes as sequences of information can be analysed. We thank an

anonymous reviewer for highlighting this issue.

3. While most methods in the timeline toolkit require a specified “start” and “end” point, the

beginning and end can be left open or vary. For several investigations, it is more important to see

if parts of the sequence overlap, rather than whether they overlap at the exact same time from the

start or before the end.

4. Details of these cases cannot be given to general public readership for obvious reasons of non-

disclosure and protecting the integrity of an investigation.

5. We appreciate that readers want to see a full case explored using these methods; however, the

central aim of this manuscript is to provide methods for academics to use to help cold case

investigations, more important than knowing methods, however, is to understand the role of an

academic in an investigation and the fact that disclosing information is not acceptable. This is also

something that academics should factor in when considering their workload – their work within

cold case investigations cannot be published.

6. Police often look for similar cases; however, researchers are able to build databases to assist with

this and provide quick-reference guides for other cases. Also, by “similar” cases, researchers can

include more historic cases that are clearly not linked to the cold case in terms of suspects, but

may be linked in terms of behaviour, modus operandi or signature.

PAGE 60j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

7. A similar CSA was conducted for another cold case: Canadian Flight 21, which was bombed mid-

flight in 1956. The bomb was placed on-board in a rear lavatory and detonated after take-off. A

crime script analysis approach outlined the most likely suspect as being one of the passengers

on-board. For more information of this analyses, please contact the corresponding author.

8. Though only publicly known information will be presented in this manuscript.

9. There are also a number of publications on sexual assault, rape and homicide cases in CSA

(Balemba and Beauregard, 2013; Beauregard et al., 2007); however, the focus here is to show

how other methods can be used to help with cold cases.

References

Bakeman, R. and Gottman, J.M. (1986), “Observing behavior: an introduction to sequential analysis”,

Cambridge University, Cambridge.

Balemba, S. and Beauregard, E. (2013), “Where and when? Examining spatiotemporal aspects of sexual

assault events”, Journal of Sexual Aggression, Vol. 19 No. 2, pp. 171-90, available at: https://doi.org/

10.1080/13552600.2012.703702

Beauregard, E., Proulx, J., Rossmo, K., Leclerc, B. and Allaire, J.-F. (2007), “Script analysis of the hunting

process of serial sex offenders”, Criminal Justice and Behavior, Vol. 34 No. 8, pp. 1069-84, available at:

https://doi.org/10.1177/0093854807300851

Briffa, M., Hardy, I.C.W., Gammell, M.P., Jennings, D.J., Clarke, D.D. and Goubalt, M. (2013), “Analysis of

animal contest data”, in Hardy, I.C.W. and Biffa, M. (Eds), Animal Contests, Cambridge University Press,

Cambridge, pp. 47-85.

Clarke, D.D. and Crossland, J. (1985), Action Systems: An Introduction to the Analysis of Complex

Behaviour, Methuen, London.

Corner, E. and Gill, P. (2019), “Psychological distress, terrorist involvement and disengagement from

terrorism: a sequence analysis approach”, Journal of Quantitative Criminology, pp. 1-28, available at:

https://doi.org/10.1007/s10940-019-09420-1

Cornish, D.B. (1994), “The procedural analysis of offending and its relevance for situational prevention”,

Crime Prevention Studies, Vol. 3, pp. 151-96.

Dawkins, R. (1976), “Hierarchical organisation”, in Bateson, P.P.G. and Hinde, R.A. (Eds), Growing Points

in Ethology, Cambridge University Press, Cambridge, pp. 7-54.

Eddy, S.R. (2004), “What is a hidden Markov model?”, Nature Biotechnology, available at: https://doi.org/

10.1038/nbt1004-1315

Ekblom, P. and Gill, M. (2016), “Rewriting the script: cross-disciplinary exploration and conceptual

consolidation of the procedural analysis of crime”, European Journal on Criminal Policy and Research,

Vol. 22 No. 2, pp. 319-39, available at: https://doi.org/10.1007/s10610-015-9291-9

Ferguson, C. (2015), “Staged homicides: an examination of common features of faked burglaries,

suicides, accidents and car accidents”, Journal of Police and Criminal Psychology, Vol. 30 No. 3,

pp. 139-57.

Ferguson, C. (2019), “Forensically aware offenders and homicide investigations: challenges,

opportunities and impacts”, Australian Journal of Forensic Sciences, Vol. 51 No. S1, pp. S128-31.

Fossi, J.J., Clarke, D.D. and Lawrence, C. (2005), “Bedroom rape: sequences of sexual behavior in

stranger assaults”, Journal of Interpersonal Violence, Vol. 20 No. 11, pp. 1444-66, available at: https://doi.

org/10.1177/0886260505278716

Gill, P., Horgan, J. and Deckert, P. (2014), “Bombing alone: tracing the motivations and antecedent

behaviors of lone-actor terrorists”, Journal of Forensic Sciences, Vol. 59 No. 2, pp. 425-35, available at:

https://doi.org/10.1111/1556-4029.12312

Homel, R., Macintyre, S. and Wortley, R. (2014), “How house burglars decide on targets”, in Leclerc, B.

and Wortley, R. (Eds), Cognition and Crime: Offender Decision Making and Script Analyses, Routledge,

Oxford, pp. 26-47.

Ivanouw, J. (2007), “Sequence analysis as a method for psychological research”, Nordic Psychology,

Vol. 59 No. 3, pp. 251-67, available at: https://doi.org/10.1027/1901-2276.59.3.251

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 61

Keatley, D.A. (2018), Pathways in Crime: An Introduction to Behaviour Sequence Analysis, Palgrave

Macmillan, London.

Keatley, D.A. and Clarke, D.D. (2019), “Matrix forecasting and behaviour sequence analysis: a combined

approach to solving cold cases”, Oral Presentation at Cold Case Investigations Evening, Murdoch

University, Perth, 23 October, pp. 1-30.

Keatley, D.A., Barsky, A.D. and Clarke, D.D. (2017), “Driving under the influence of alcohol: a sequence

analysis approach”, Psychology, Crime and Law, Vol. 23 No. 2, pp. 135-46, available at: https://doi.org/

10.1080/1068316X.2016.1228933

Keatley, D.A., McGurk, S. and Allely, C.S. (2019), “Understanding school shootings with crime script

analysis”, Deviant Behavior, pp. 1-13.

Keatley, D.A., Marono, A. and Clarke, D.D. (2018), “Unmaking a murderer: behaviour sequence analysis

of false confessions”, Psychiatry, Psychology and Law, Vol. 25 No. 3, pp. 425-36.

Keatley, D.A., Golightly, H., Shephard, R., Yaksic, E. and Reid, S. (2018), “Using behavior sequence

analysis to map serial killers’ life histories”, Journal of Interpersonal Violence, available at: https://doi.org/

10.1177/0886260518759655

Kempe, D., Kleinberg, J. and Tardos, E. (2003), “Maximizing the spread of influence through a social

network”, Proceedings of the Ninth ACM SIGKDD International Conference of KNOWLEDGE Discovery

and Data Mining, pp. 137-46.

Krebs, V.E. (2002), “Mapping networks of terrorist cells”, Connections, Vol. 24 No. 3, pp. 43-52, available

at: https://doi.org/10.1.1.16.2612

Lawrence, C., Fossi, J. and Clarke, D. (2010), “A sequential examination of offenders’ verbal strategies

during stranger rapes: the influence of location”, Psychology, Crime & Law, Vol. 16 No. 5, pp. 381-400,

available at: https://doi.org/10.1080/10683160902754964

Leclerc, B. and Wortley, R. (2014), Cognition and Crime: Offender Decision Making and Script Analyses,

Routledge, Oxford, available at: https://doi.org/10.4324/9780203083482

Leclerc, B., Wortley, R. and Smallbone, S. (2011), “Getting into the script of adult child sex offenders and

mapping out situational prevention measures”, Journal of Research in Crime and Delinquency, Vol. 48

No. 2, pp. 209-37, available at: https://doi.org/10.1177/0022427810391540

Magnusson, M.S. (2000), “Discovering hidden time patterns in behavior: T-patterns and their detection”,

Behavior Research Methods, Instruments, & Computers: A Journal of the Psychonomic Society, Inc, Vol.

32 No. 1, pp. 93-110, available at: https://doi.org/10.3758/Bf03200792

Magnusson, M.S., Burgoon, J.K. and Casarrubea, M. (2016), Discovering Hidden Temporal Patterns in

Behavior and Interaction, Springer, New York, NY.

Marono, A., Clarke, D., Navarro, J. and Keatley, D. (2018), “A sequence analysis of nonverbal behaviour

and deception”, Journal of Police and Criminal Psychology, Vol. 33 No. 2, available at: https://doi.org/

10.1007/s11896-017-9238-9

Marono, A., Clarke, D.D., Navarro, J. and Keatley, D.A. (2017), “A behaviour sequence analysis of

nonverbal communication and deceit in different personality clusters”, Psychiatry, Psychology and Law,

Vol. 24 No. 5, pp. 730-44.

Miller, G.A. and Selfridge, J.A. (1950), “Verbal content and the recall of meaningful material”, Journal of

Psychology, Vol. 63 No. 2, pp. 176-85.

Roach, J. (2019), “The retrospective detective: cognitive bias and the cold case investigation”, Decision

Making in Police Enquiries and Critical Incidents, Palgrave Macmillan, London, pp. 129-50, available at:

https://doi.org/10.1057/978-1-349-95847-4_8

Taylor, O., Keatley, D.A. and Clarke, D.D. (2017), “A behavior sequence analysis of perceptions of

alcohol-related violence surrounding drinking establishments”, Journal of Interpersonal Violence,

available at: https://journals.sagepub.com/doi/abs/10.1177/0886260517702490?journalCode=

jiva#articleCitationDownloadContainer

Taylor, P.J. (2006), “Proximity coefficients as a measure of interrelationships in sequences of behavior”,

Behavior Research Methods, Vol. 38 No. 1, pp. 42-50.

Taylor, P.J., Jacques, K., Giebels, E., Levine, M., Best, R., Winter, J. and Rossi, G. (2008), “Analysing

forensic processes: taking time into account”, Issues in Forensic Psychology, Vol. 8 No. 8, pp. 43-56.

PAGE 62j JOURNAL OF CRIMINAL PSYCHOLOGYjVOL. 10 NO. 2 2020

Townsend, E., Wadman, R., Sayal, K., Armstrong, M., Harroe, C., Majumder, P. and Clarke, D.D. (2016),

“Uncovering key patterns in self-harm in adolescents: sequence analysis using the Card Sort Task for

Self-harm (CaTS)”, Journal of Affective Disorders, Vol. 206, pp. 161-8, available at: https://doi.org/

10.1016/j.jad.2016.07.004

Wadman, R., Clarke, D., Sayal, K., Armstrong, M., Harroe, C., Majumder, P., Vostanis, P. and Townsend,

E. (2017), “A sequence analysis of patterns in self-harm in young people with and without experience of

being looked after in care”, British Journal of Clinical Psychology, Vol. 56 No. 4, pp. 388-407.

Wortley, R. (2016), “Situational precipitators of crime”, Environmental Criminology and Crime Analysis,

2nd ed., available at: https://doi.org/10.4324/9781315709826

Zucchini, W., MacDonald, I.L. and Langrock, R. (2017), Hidden Markov Models for Time Series: An

Introduction using R. Chapman and Hall/CRC.

Further reading

Keatley, D.A., Arntfield, M., Gill, P., Clare, J., Oatley, G., Bouhana, N. and Clarke, D.D. (n.d.), “Behaviour

tracking: using geospatial and behaviour sequence analysis to map crime”, Security Journal.

Corresponding author

David Keatley can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website: www.emeraldgrouppublishing.com/licensing/reprints.htm Or contact us for further details: [email protected]

VOL. 10 NO. 2 2020 j JOURNAL OF CRIMINAL PSYCHOLOGYj PAGE 63