make summary 5 pages (((this book Sociocultural Behavior Sensemaking: State of t

Question

make summary 5 pages

(((this book

Sociocultural Behavior Sensemaking: State of the Art in Understanding the Operational Environment Edited by: Jill Egeth, PhD Gary L. Klein, PhD Dylan Schmorrow, PhD)))

Section Two: Detecting

Detecting sociocultural factors and elements in an environment pg. 107

Lashon Booker

5 Transforming data into information: Enabling detection and pg. 111

discovery for sociocultural analysis John M. Irvine

6 Current trends in the detection of sociocultural signatures: pg. 147

Data-driven models Antonio Sanfilippo, Eric Bell & Courtney Corley

7 Visualization for sociocultural signature detection pg. 173

Ronald D. Fricker, Jr., Samuel E. Buttrey & William Evans

8 Cross-cultural training and education for detection pg. 217

Sharon Glazer, Lelyn Saner, Ivica Pavisic & Molly Barne

Explanation / Answer

Detecting sociocultural factors

Capability Area 2: Detecting
“Capabilities to discover, distinguish, and locate operationally relevant sociocultural signatures
through the collection, processing, and analysis of sociocultural behavior data” (Schmorrow, 2011,
p. 42).
Sociocultural signatures that potentially have operational relevance cover a vast domain,
encompassing the perceptions, sentiments, attitudes, and behaviors of various populations of
interest. Challenges in developing a robust capability to detect these signatures are equally
daunting. Such a capability depends on processing and then analyzing voluminous collections of
relevant data sources, with data types ranging from raw sensor measurements to text, image, and
audio data, to qualitative assessments about attitudes. The data cover varying time spans, and
describe phenomena at a variety of scales and levels of detail. Further, detection requires analysts
to resolve many technical issues, including the need to isolate meaningful signals in a deluge of
noise, manage both structured and unstructured data, process data streams in real time, and
identify rare events. The chapters in this section describe state-of-the-art capabilities available to
meet some of these challenges, as well as emerging capabilities that are likely to become useful in
the near future.
Data Processing
Irvine’s chapter, “Transforming data into information: enabling detection and discovery for
sociocultural analysis,” reviews the issues involved in transforming raw data into the information
that researchers need to detect interesting sociocultural patterns. The discussion focuses primarily
on four types of data sources: surveys, social media, imagery, and video. Surveys provide controlled
methods for collecting data about individuals and societies. Social media supply direct and timely
data about the opinions of individuals, as well as the unfolding of events. Imagery is becoming an
increasingly valuable source of data about societies in general as well as local attitudes and
behaviors. Video contains data that can be used to analyze temporal events and detect both simple
activities and complex behaviors.

Computational Modeling
In “Current trends in the detection of sociocultural signatures: data-driven models,” Sanfilippo,
Bell, and Corley address detection of sociocultural activities and events from a modeling
perspective. The prominent theme here is that analysts must “harvest” data in forms that enable
them to build, calibrate, and run computational models of sociocultural phenomena. This means
that parameters of data records relevant to the content categories of interest must first be
extracted and measured to produce sociocultural data signatures. Many techniques are available to
assist in this process. Computational models then use the data signatures to characterize the
behavioral patterns of interest in the underlying data. While there are many different kinds of
computational models, all can be viewed in terms of sociocultural model signatures that specify
how to detect and assess sociocultural patterns.
Social media content provides some unique challenges and opportunities for computational
modeling. The authors review new data harvesting methods specifically designed to address those
challenges, along with examples of the kinds of models that can use social media data to detect
sociocultural patterns. They also provide a brief discussion about how detection models can be
used as a starting point to help make predictions.
Visualization

Transforming data into information

Nowadays it seems impossible to open any recent issue of a technical journal, trade newsletter, or
popular science publication that does not contain at least one article about “big data.” To be sure,
digital data is accumulating at a staggering rate – roughly 80 petabytes an hour, more than 10
times larger than the holdings of the Library of Congress (Smolan & Erwitt, 2012). The combination
of systematic data collection by commercial and government organizations, the emergence of
unstructured digital data sources from news and social media, and advances in sensor technology
has led to this explosion of data. For researchers seeking to detect interesting sociocultural
patterns, the volume of data presents both opportunities and challenges. The opportunities stem
from the availability of data to describe many aspects of the daily life of individuals and societies.
The challenge is distilling this massive quantity of bits and bytes into meaningful information that
supports deeper analysis.
In approaching the derivation of understanding from raw data, we present a conceptual framework
that considers the basic processing components (Figure 1). The raw data can originate from any
number of sources: surveys, social media, news reports, and sensor data. One can exploit each of
the sources individually, but greater benefits often arise from synergistic analysis of multiple
sources, i.e., information fusion (Hall, 1992; Klein, 1999). The timeline for developing inferences
from the data depends on the specific mission or application. In some tactical settings users may
need to derive inferences, including warnings or alerts, under demanding time constraints. Deeper
analysis, including forensics, might require days, weeks, or even months. The ways in which users
might interact with the analytic tools differ for these two types of missions.
1.1. Data Sources
Many sources of data provide useful information about sociocultural issues, either directly or
indirectly. Traditional tools for social science research have included direct observations and
surveys, as well as published media such as news reports. More recently, social media have
changed the nature of the data, because now the actors in a society provide information directly
1.2. Sources to Evidence
Raw data generally have only limited value. Some processing is necessary to extract useful
information from the original sources. The methods for transforming sources of raw data into
information or evidence depend on both the nature of the source data and the information
required by the end user. For example, processing the results of a survey to assess support for a
political candidate can be simple. The tabulation of the responses gives an initial indication of
findings, and more sophisticated processing would weight the responses in accordance with the
sample design. Still more sophisticated analysis might attempt to assess the veracity of responses
or impute values for missing responses based on answers to related questions (Cooley & Lohnes,
1971). More challenging data sources include unstructured and semi-structured text. For example,
a researcher might want to explore the level of support in neighboring Arab states for the Syrian
rebels seeking to overthrow President Assad by examining posts on social media sites. Turning the
text data into relevant information involves performing sentiment analysis on relatively
unstructured text that could include slang and country-specific references. This type of analysis
presents far more challenges and the tools available today, while promising, are less than perfect.

1.3. Evidence to Inference
Armed with useful information, researchers must next draw inferences based on the data. Various
inferences are possible, depending on the research questions and the nature of the data. Common
types of inference include:
Discovery: Uncovering a new phenomenon or relationship. The researcher has access to
information that has not been fully exploited and should have some working hypotheses, but the
exploration of the data could reinforce or refute these hypotheses. More important, the
exploration can suggest new relationships to consider. Data mining offers a set of tools commonly
used for discovery.
1.4. User Understanding
Researchers could perform the preceding steps in processing via automated methods, user
manipulation of the data, or a combination of human and automated processing. The final step
requires a human to interpret the results of the processing. Deeper analysis and understanding can
only come when the user examines the inferences produced by processing, displaying, and
visualizing the information, and explores the relationships. A separate chapter in the Detecting
section of this book explores data visualization in more depth. For the purposes of the present
discussion, however, it is important to realize that users need tools and methods for querying the
data, displaying information, and understanding the data provenance.

2. Data Sources and Information
Almost any source of data contains information, either direct or indirect, about the society and
culture that produced the data. Certain data sources, however, have emerged as particularly useful
for understanding sociocultural phenomena. In this chapter, we focus on four specific sources:
survey data, social media, imagery data, and video data. Some of the data sources we discuss are
also addressed in other chapters in this volume; in the present chapter we focus on the use of data
and information for purposes directly related to detection.
2.1. Survey Data: Controlled Methods for Data Collection
A basic approach to understanding individuals and societies consists of asking them questions
directly. A conversation with one individual is termed an interview. While interviews can be
revealing, the information is specific to the person being interviewed. A broader discussion with
multiple participants, in the form of a focus group, offers a systematic way to elicit a broader set of
information. Extending the concept to a structured approach that elicits information from many
respondents moves the process into the realm of survey research. Depending on the research
questions, a trade-off exists between the rich narrative available from open-ended questions with
free-text responses and specific, focused questions amenable to quantitative analysis.

2.2. Social Media: Text and Multimedia Data
The use of social media has exploded in recent years, making multiple sources of data available to
anyone who can harvest all the bits and bytes. Analysis of social media data appeals for several
reasons: it provides direct access to people’s comments and opinions, the information is readily
available, and very little time elapses between the unfolding of an event and the corresponding
reactions in social media. Because social media offer direct links to events as they happen,
numerous researchers are exploring ways to leverage this data for understanding a wide range of
geopolitical, economic, and sociocultural issues.

Current trends in the detection of sociocultural

1. Introduction
Challenges to the security, health, and sustainable growth of our society keep escalating
asymmetrically due to the accelerating pace of global change. The increasing velocity and volume
of information sharing, social networking, economic forces, and environmental change have
expanded the number and frequency of “game-changing moments” that a community can face.
Now more than ever, we need anticipatory reasoning technologies based on sociocultural
understanding to detect, analyze, and forecast potential change so that we can plan appropriate
interventions to neutralize adversaries and protect the public (Costa & Boiney, 2012). The creation
of such a “social radar” starts with the detection of sociocultural signatures in data streams.
By harvesting behavioral data and analyzing them through evidence-based reasoning, we can
detect sociocultural signatures in their context to support situation awareness and decision
making. Developers use the harvested data as training materials from which to infer computational
models of sociocultural behaviors or calibrate parameters for such models. Harvested data also
serve as evidence input that the models use to generate insights about observed and future
behaviors for targets of interest. This input often results from assembling data of diverse types and
aggregating them into a form suitable for analysis.
2. Current Approaches to Sociocultural Signature Detection
This section defines sociocultural signatures and reviews existing methods for detecting them. For
expository purposes, we use the modeling of sociopolitical contention and violent intent as the
case study, but the methods described apply to other domains as well.
We distinguish between sociocultural data signatures (SDSs) and sociocultural model signatures
(SMSs). An SDS can be envisioned as a set of attribute-value pairs describing a data record, as
defined in (1), and exemplified in (2) with reference to violent intent (vi), where attributes denote
classes of words (e.g., military = {war, soldier, weapon…}) and the associated values indicate how
frequently the instances of the attribute occur in a given message.

2.1. Acquisition of SDSs
Content analysis is perhaps the most widely used technique to distill SDSs from data that include
surveys, interviews, ethnographies, social media, news wires, and public speeches. This
methodology interprets text through categorical annotation to study the content of
communication (Holsti, 1969; Krippendorff, 2004). For example, assessing violent intent through
content analysis involves identifying categories of meaning correlated with the expression of
violent behavior or the lack thereof. Analysts then draw inferences from the occurrence of such
categories in the document(s) reviewed to estimate the likelihood that the communication source
would engage in violent behavior.

Visualization for sociocultural signature detection

1. Introduction
Data visualization can reveal relationships that summary statistics simply cannot convey. The
canonical example is Anscombe’s data, plotted in Figure 1 (Anscombe, 1973). Visually these four
sets of data clearly show very different relationships between the x and y variables, yet the means
and standard deviations of each of the x variables are exactly the same; the means and standard
deviations of the y variables are also the same; the correlations between x and y in each of the four
cases are the same; and, even the regression fits are the same. Merely examining some summary
statistics without plotting the data could completely mislead a viewer into believing that there is
little difference in the underlying phenomena.

1.1. Visualization for Sociocultural Signature Detection
As defined in the Office of the Secretary of Defense (OSD) Human Social Culture Behavior (HSCB)
Modeling program, sociocultural signature detection is one of the four program capabilities:
Understand, Detect, Forecast, and Mitigate. The Detect capability consists of “capabilities to
discover, distinguish, and locate operationally relevant sociocultural signatures through the
collection, processing, and analysis of sociocultural behavior data” (Office of the Secretary of
Defense, 2013). The site goes on to say,
Once the defining features of the sociocultural setting are understood, the next steps are to
develop a persistent capability to detect sociocultural behavior signals of interest amidst
complexity and noise, and to harvest data for analysis. This entails capabilities for ISR in the
area of sociocultural behavior (referred to here as a “social radar”), with particular focus on the
challenges associated with open source data collection. It also requires robust systems for
storing and managing that data, and tools enabling timely, dynamic analysis.
Visualization, then, supports “enabling timely, dynamic analysis” in terms of both finding relevant
signatures and identifying when known signatures change. The former is largely a retrospective
exercise in exploring and modeling existing data to identify sociocultural signatures, while the latter
is a prospective exercise in monitoring a given signature to identify if and when it changes. Both
types of analysis may require a variety of cross-sectional, temporal, and spatiotemporal analytical
methods and visualization techniques for infrastructure, social, and other types of network data;
Geographic Information System (GIS) and similar types of spatial data; surveys and related types of
data; and social media and other types of linguistic data.4
Good visualization for sociocultural signature detection must be optimized for the particular
application and user. For example, the best methods for visualizing networks may be quite
different than those for displaying GIS data. Similarly, GIS methods for displaying point data on
maps are not appropriate for displaying areal data from surveys. Some types of social media data.

1.2. Visualization as Part of Data Exploration
Exploratory data analysis or EDA (Tukey, 1977) focuses on summarizing data to make it easy to
understand, often through graphics and data displays, and generally without using formal statistical
models or hypotheses. As described by the National Institute of Standards and Technology (NIST,
2012):
Exploratory Data Analysis (EDA) is an approach/philosophy for data analysis that employs a
variety of techniques (mostly graphical) to
• maximize insight into a data set;
• uncover underlying structure;
• extract important variables;
• detect outliers and anomalies;
• test underlying assumptions;
• develop parsimonious models; and
• determine optimal factor settings.

1.3. Caution: Apophenia and Pareidolia
Shermer (2008) noted that, “...our brains are belief engines: evolved pattern-recognition machines
that connect the dots and create meaning out of the patterns that we think we see in nature.
Sometimes A really is connected to B; sometimes it is not.” Thus, while the human brain excels at
finding true patterns against noisy backgrounds, it also quite capable of over-interpreting pure
noise to find nonexistent “patterns.” This phenomenon of finding (seemingly) meaningful patterns
in random or meaningless data is called apophenia; pareidolia is visual apophenia (Hoopes, 2011).

2. Visualization to Detect Sociocultural Signatures
In this section we first review some of the classical methods and approaches for displaying data.
We then move on to discuss visualization methods and techniques in the context of fairly general
problem and data classes — networks, geographically based data, surveys, linguistic data, and
social media — that are particularly relevant to sociocultural signature detection.
2.1. Classical Approaches to Data Visualization
Data visualization has historically been thought of in terms of graphing numerical data. Typically
these data were either cross-sectional (i.e., collected at the same point in time) or longitudinal (i.e.,
collected over time) and were either continuous or discrete, with different methods developed to
address each combination. For example, typical graphical methods for continuous cross-sectional
data include histograms and box plots, while methods for discrete cross-sectional data include a
variety of bar and pie charts, as well as simple tabular summaries. Longitudinal data are most
typically plotted on some type of time series plot.
Visualization methods for two variables include scatterplots, mosaic plots, side-by-side box plots,
and others. The scatterplot matrix (also known as a pairs plot) permits simultaneous presentation
of multiple scatterplots. It is typically displayed in a square showing n continuous variables on both
the vertical and horizontal axes, where each row and each column depicts one variable’s
comparison with the other n 1 variables. The diagonal is typically left for labels or single-variable
2.2. Network Visualization
Networks occur naturally in many situations. Conceptually they are quite simple, with every entity
denoted by a node in the network and linkages between entities denoted by arcs. Visualizing
networks in ways useful for learning about and understanding the network, on the other hand, can
be anything but simple. For example, particularly with large networks, simply displaying all the arcs
and nodes results in the classic “hairball” (see, for example, Figure 7) from which users can discern
2.3. Visualization of Geographic Information
The visualization of geographic data – the making of maps – has a long history and recent
developments in computing have put the display of geographic data within easy reach.
Furthermore, the widespread incorporation of Global Positioning System (GPS) capabilities in
objects such as mobile phones and freight palettes has produced significant growth in the amount
of georeferenced data available to consumers and organizations. Additionally, it is now
straightforward to perform analyses, as distinct from visualizations, that would have been much
more difficult even ten years ago. By analysis, we refer here to tasks such as identifying the
polygons that contain particular points (for example, identifying terrorist incidents with provinces),
projecting points onto lines (for example, identifying traffic accidents with road locations), or
computing lines of sight (for example, locating regions that cannot be seen from a particular
tower). Of course, the distinction between display and analysis is not always sharp.
2.3.1. Coordinate systems
One aspect that sets geographic visualization apart from traditional approaches is that developers
must choose the coordinate systems with care. At the start, data are often associated with
geographic coordinates—latitude and longitude, typically in degrees. Since the earth is roughly
ellipsoidal, geographic coordinates must be associated with a “datum” that characterizes the
particular ellipsoidal approximation in use. A latitude, longitude pair for a particular location
derived under one datum might be hundreds of meters away from the same pair derived under
another datum. A number of datums7 are available, but the WGS84 datum, used by GPS, is by far
the most common. Several computer programs make it possible to convert coordinates based on
one datum to another.
Over small areas geographic coordinates can be plotted directly, neglecting the curvature of the
earth, but for even moderate distances (say, dozens of miles or scores of kilometers) a projection is
necessary. As one interesting example, the two towers of New York’s Verrazano-Narrows bridge
are 1-5/8 inches farther apart at the top than at the bottom because of the curvature over the
bridge’s 4,260-foot length.

Cross-cultural training and education for detection

1. Introduction and Overview
Sociocultural signatures are unique, identifiable (often because they are repeated) features of the
social and cultural landscape. That landscape includes observable individual biometrics (i.e.,
characteristics or traits, such as personality profiles, fingerprints, and voice profile), as well as
individual and group sentiments and behaviors (Maybury, 2010) exhibited in political, economic,
and social structures.
Detection of sociocultural signatures involves more than maintaining situation awareness and
identifying facts in the physical environment. It requires a deeper understanding of what those
signatures mean. Sociocultural analysis helps to unravel the meanings of these signatures by
decoding how and why people sense the world as they do.
2. What Is Culture?
Culture gives meaning and distinction to the existence of a group. The term “culture” encompasses
values, norms, behaviors, and beliefs implicitly shared among members of a social system – defined
as a group characterized by meaningful interactions among individuals (Schwartz, 2009). These
characteristics of a culture develop and modify over time, and are reinforced by the interactions of
people within it. Thus, an inherent component of understanding culture is understanding how the
people within it interact and why. For this reason, intelligence professionals (IPs—a broad term
that includes analysts) must focus on patterns of social interactions that provide evidence of group
behaviors and how behaviors and sentiments develop. Starting with dyadic relationships, including
person-to-person, person-to-group, and group-to-group, IPs can come to understand the more
complex relationship structures that define cultures, and explain why dyadic relationships take a
particular form.
The term “culture” is multi-layered and can be applied to describe different kinds of social systems,
including regions, countries, nations, ethnic groups, and families. While many associate the concept
with nationality, not all social systems correspond to national borders (McGinn, Weaver,
McDonald, van Driel, & Hancock, 2008). Thus, we can refer to peer group cultures, corporate
cultures, and national cultures.
Culture influences and is influenced by social, educational, business, political, economic, linguistic,
legal, and religious systems (Tayeb, 1994). It is both around and within us, and individuals have
cultural signatures, but people are not embodiments of culture. By analogy, simply because the
United States is considered a wealthy nation does not mean that all people in the United States are
wealthy. Similarly, even within a relatively collectivistic culture different individuals focus to
different degrees on their duty to their families and subgroups.
2.1. Culture-General Concepts
Culture-general concepts central to CCTE include artifacts and practices, norms, beliefs, and values.
Artifacts and Practices. Artifacts are aspects of culture that are immediately visible; they include
tangible objects or observable practices. Practices are behaviors or patterns of social interactions,
which in turn reflect an underlying set of rules and understandings. A group’s practices are
informed by the group’s values and norms. Because of this, a practice may carry different
implications across groups. For example, when a subordinate speaks to a supervisor, it is common
practice in U.S. culture that the two maintain eye contact, as this implies attentiveness and
sincerity (Hattori, 1987). In Japanese culture, however, a subordinate making significant eye
contact with a supervisor is considered disrespectful or immodest (Hattori, 1987)—likely due to
how highly Japanese society values hierarchy (Schwartz, 1999).
Considered together, artifacts and practices form “surface culture.” After detecting artifacts and
practices, IPs must consider how they came to be, which addresses “process culture,” and why,
which addresses “deep culture.” Norms, beliefs, and values represent both process and deep
cultures.

3. Science and Technology Gaps in Cross-Cultural Training and Education
CCTE can strengthen intelligence professionals’ (IPs’) abilities to detect anomalies in foreign others’
interactions and behaviors, and to explain the assumptions underlying them in order to forecast
their implications for security and to mitigate threat. However, we have observed that IPs perceive
themselves to have limited autonomy in preparing statements that consider alternative
perspectives that might explain observed artifacts.
Indeed, the most prominent gap in the IP’s toolkit concerns the ability to answer “how” and “why”
observed cultural factors emerged. In many cases, IPs who must answer an intelligence question
cannot explain how and why the answers are manifested as they are. In some cases, they may not
have the opportunity because the answer to why is not part of the intelligence question posed. Too
often, however, IPs explain artifacts on the basis of their own cultural understanding of an
adversary’s behaviors, rather than being able to present why the behavior occurs as it does from
the perspective of the adversary. For example, in a recent NBC News report, Windrem (2013)
reported a cyber analyst’s observation that different Chinese groups engaging in cyber-hacking
activities do not cooperate with each other because they do not share information. A close
examination of the explanation suggests a lack of cultural understanding: the lack of sharing is
indicative of not cooperating, but does not explain it. CCTE would help an IP recognize that this lack
of intergroup cooperation is a manifestation of the cultural values for mastery and hierarchy.
3.1. Self-Awareness
One of the shortcomings that IPs must overcome is minimal awareness of their own cultural biases
when interpreting others’ behavior. Attribution bias—in which people explain their own
experiences as a function of the environment, but others’ behaviors as a function of their
personality or personal attributes—might lead IPs to misattribute explanations to personality
instead of situation and culture. Thus, the first step in enhancing intercultural interactions within
the intelligence community consists of learning about or reflecting upon our own character and
past experiences.
Self-awareness inventories enable people to explore their own thinking patterns and behavioral
styles. These tools can serve as a springboard for thinking about how those patterns bias the
interpretations of artifacts. They also provide guidance for trainers who need to understand the
stage that trainees have reached in their development. Objectives associated with self-assessment
are: (a) provide instrumented feedback to individuals regarding their self-construals, values, and
level of intercultural competence, (b) introduce training concepts, (c) supply non-threatening
vocabulary, (d) serve as a frame of reference, (e) teach trainees to appreciate the strengths and
understand the limitations of people different from themselves, and (f) help individuals explore
ways to improve their effectiveness as senders or receivers of intercultural communication. At the
end of a self-awareness training module, participants should better understand themselves and
have a general sense of the areas in which they must improve to enhance the quality of their
intercultural interactions. They will also begin to understand the constraints placed on them when
engaging in intercultural interactions.
3.2. Perspective-Taking
The next step in the CCTE process is learning to take the other’s(Heuer & Pherson, 2011) and “red teaming” (Green Sands & Haines, 2013; Mateski, n.d.). The goal
is not to provide sophisticated stereotypical explanations that label cultures as falling within one
category of a single cultural dimension (e.g., either individualistic or collectivistic), but to untangle
the complex web of cultural syndromes that together create a unique cultural explanation within a
particular situation and context (Osland, Bird, DeLano, & Jacob, 2000).
3.3. Surface, Process, and Deep Culture
Once IPs understand the complexities of cultural syndromes and individual profiles, they should
learn how to identify artifacts or indicators that together could illuminate a cultural signature. For
example, an IP might observe a paved highway that ends before it connects to another roadway.
The IP must then explore what the highway symbolizes to the community and what its completion
would mean to the people in the cultural setting. The IP might discover that the community views
the incomplete road as a government ploy to show citizens that it is working toward fulfilling its
promises, but at the same time needs the community to contribute to the completion of the road
by supporting the re-election of the ruling party. Then the IP can begin to probe the layers of
assumptions and underlying belief structures. Perhaps there is a generalized pattern of societal
cynicism, exemplified by beliefs that people “will stop working hard after they secure a
comfortable life” and “powerful people tend to exploit others” (Leung et al., p. 293). The
underlying cultural values that lead to this generalized belief include hierarchy and mastery—
principles that guide status and rules for dominating others. The IP must therefore peel back the
surface layer of artifacts, the middle or process layer of meaning, and the deep core of assumptions
to understand a culture.
3.4. Increasing Awareness of Communication Patterns
When observing dyadic interpersonal communication, IPs must first have a good model of the
observable elements that identifies if and how the elements are connected: who communicates
with whom; what, when, and how often they communicate; and what types of information are
exchanged. IPs can apply social network analysis to gain this insight. A second fundamental aspect
of communication concerns the role that each individual plays in the communication network (e.g.,
primarily a sender or a receiver). Role theory can help to illuminate this aspect. Finally, IPs must
consider the detailed characteristics of the communication that occurs, including the semantics,
grammar, syntax, and other nonverbal factors that together can yield a cultural signature.
Together, these components form the basis for understanding relationship structures.
4. State-of-the-Art
This section describes state-of-the-art findings on perspective taking and communication patterns
that inform CCTE designs.
Researchers have made numerous advances in the past decade that assist people to take on
different cultural perspectives and improve understanding of relationship structures (Dien, Blok, &
Glazer, 2011). These findings have yet to be incorporated in CCTE, as the protocols for perspectivestructures as tools for detecting and explaining identified artifacts. CCTE programs should
incorporate findings from these research streams to the extent they are relevant to each agency’s
CCTE goals.
4.1. Perspective-Taking
One of the enabling objectives of CCTE is to prepare IPs to consider alternative perspectives.
Current research from laboratory studies offers promising options for CCTE. Researchers have
demonstrated methods that cue people to view artifacts through a cultural lens (e.g., individualism
or collectivism) different from their own (e.g., Gardner, Gabriel, & Lee, 1999; Han, 2010; Oyserman
& Lee, 2008a; Oyserman & Sorensen, 2009). Cultural priming studies (e.g., Gardner et al., 1999;
Han, 2010; Oyserman & Lee, 2008b; Oyserman, Sorensen, Reber, & Chen, 2009) have provided
behavioral evidence through application of psychological and neuroscientific methodology that
people from North America and East Asia can be cued to think from different cultural perspectives,
as evidenced by participants’ change in endorsed values (Briley & Wyer, 2001, 2002; Gardner et al.,
1999; Yang & Bond, 1980). Several studies have also examined the effects of priming on judgments
about specific scenarios, such as acceptance of euthanasia or affirmative action (Kemmelmeier,
2003; Kemmelmeier, Wieczorkowska, Erb, & Burnstein, 2002).
4.2. Communication Patterns
Network Analysis. CCTE programs can employ a number of tools and techniques for network
analysis: the process of modeling the relationships among a set of individuals, groups, or other
entities based on data about them (Heuer & Pherson, 2011). Each entity is a node in the network;
analysis identifies and quantifies the relationships, or linkages, between the nodes. These
techniques are very useful for analysis of dyadic communications in that each data point represents
a connection between two entities (e.g., observing the number of times Bill initiated a
communication with George). By this measure, the strength of the link between Bill and George is a
function of the total number of communications exchanged between them. It is critical, then, for
IPs to be aware of the measurement unit used to determine associations (e.g., number of
communications) when interpreting a network model.

4.3. Cross-Cultural Training and Education
People in the United States tend to view training and education as the ‘culturally appropriate’
processes for developing work-related competencies, including competencies for international
assignments (Fowler, 1994). Although cultural immersion programs, where a person embeds in a
society for an extended period of time, can also be helpful, such programs tend to send people to
one country only and can be quite costly. In this section, we address the state-of-the-art in CCTE,
including several different types and modes of training that have proven useful, such as critical
incidents, role-plays, and scenario-based training.
Since the 1990s, much research has been devoted to demonstrating the benefits of CCTE, including
cost reduction, improved performance, and more effective decision making (Brugman, Reinhart,
Feinberg, Glazer, Falk, & Castle, 2010; Brugman, Reinhart, Feinberg, Falk, & Castle, 2012). CCTE
teaches people to engage in deeper analysis of surface observations. Knowledge and skills acquired
from CCTE are important for creating cultural fluency and better understanding of others. These
competencies give people tools to attend to the big picture, encourage cognitive flexibility, and
reinforce a proclivity for asking searching questions.
4.4. Summary
In this section, we outlined some findings from state-of-the-art research on culture, as well as
modes and media for CCTE. An important point to make here is that we do not advocate any one
particular mode or teaching medium for CCTE, because training must be tailored to the audience
and need. We do, however, endorse cross-cultural (culture-general) training as a foundation for all
DoD and USG personnel in order to strengthen cognitive flexibility, openness, and ability to engage
in sociocultural analysis.
5. Transitioning Cross-Cultural Training and Education into Operations
CCTE can contribute to developing capabilities to discover, distinguish, and locate operationally
relevant sociocultural signatures derived from sociocultural behavior data. Although not everyone
has the opportunity to gain experience through international contacts, CCTE can help IPs acquire
knowledge and skills in detecting cultural nuances. Most CCTE programs focus on preparing
sojourners to interact with people in different cultures. However, in this type of CCTE, the goal is
not necessarily to prepare students for physical interactions, but to enable them to detect cultural
underpinnings in extracted excerpts of communications. The activities in which trainees engage
during CCTE programs can also be applied to this type of task.
We suggest that designers of CCTE programs roll out the curriculum in stages, corresponding to the
aspects discussed earlier. First, IPs would receive cultural awareness training in which they learn
5.1. Recommendations for Transitioning CCTE Activities to Operations
Scenario-Based Training. SBT appears to be one of the most cost-effective ways of implementing
CCTE, but multiple factors influence the success of a particular SBT. These factors include the
expertise of the trainer, the organization’s reward systems, the climate for error management, the
learning environment, task requirements, the student’s ability to transfer learning to real life,
individual motivation, formative and summative evaluation strategies, content of the training
program (including method, strategy, and tools), student’s real-work life experience (i.e.,
accumulated time on the job), and the student’s intellect (Salas, Priest, Wilson, & Burke, 2006).
Developing an SBT is time consuming and costly upfront, but in the long run the benefits outweigh
the costs, as the program can be used with all trainees. The upfront costs cover selecting and
populating the relevant scenarios (e.g., pulling archival intelligence that might be relevant to the
Boston Marathon bombers) from which designers can identify the competencies necessary for
students to perform effectively on the job, as well as any deficiencies in performance that require
correction. If the scenario targets the right deficiencies, designing and implementing SBTs that
focus on the upgraded competencies becomes easier, as does evaluating the training’s success.
Furthermore, once core competencies have been established, trainers can generate corpora of
scenarios and assessment instruments for each, and then store, modify, update, and reuse them.
5.2. CCTE Learning Sequence
Drawing upon practices that trainers use with sojourners, such as training pre-departure and postarrival,
we also recommend that operational organizations use training as an introduction, and also
provide continuous learning and reinforcement to sustain IPs’ abilities to adopt different cultural
perspectives. Ideally, this would occur by default in immersion experiences, but our
recommendations apply in the cases where those opportunities are not available or cost effective.
As with pre-departure training, a goal is to establish learners’ expectations about working with
cultural materials. Such training would reduce students’ anxieties, increase their confidence that
they can detect relevant cultural information, and reassure them that all information observed is
culture laden. Similar to post-arrival training, on-the-job reinforcement of culture studies would
address real-time issues as experienced in the actual work setting. This is all the more important as
IPs can become so immersed in their subject matter that they are sometimes unable to distinguish
between their own cultural biases and those of the culture they are analyzing.
5.3. Evaluating Operational Utility
Empirical inquiry has just begun to explore the extent to which cross-cultural training has proven
useful for IPs or military personnel deployed abroad. Criteria for performance measurement are
still in development, and may not be available for a few more years. Even so, organizations have
ways to determine the utility of CCTE programs. One indicator of CCTE success would be a
noticeable increase in the depth or richness with which IPs interpret a situation. This measure
would require a comparison of baseline analytic performance to performance after CCTE.
Instructors might evaluate the utility of CCTE programs across multidisciplinary and multi-agency
teams of IPs, including technology experts and social scientists.
We also strongly recommend that members of the analytic community work with people outside
their own groups or agencies. The sharing of expertise between people from different groups and
agencies can lead to insights and perspectives across disciplines—a process very much akin to
detecting cultural signatures in the world at large.

Related Questions

Navigate

• Browse (All)
• Browse M
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.