Development of Clinical Discourse Analysis

1.     Developed while working in the Albuquerque Public Schools (1976-1978)
2.     Recognized that the contemporary tests and sampling procedures were poor
3.     Determined that more attention to function rather than superficial aspects of
            language was needed.
4.     Initiated a data-grounded and interpretive Research Project
5.     Created a set of criteria to identify individuals who appeared to exhibit language
         difficulties.  Had to meet three criteria:
               Diagnosed as language disordered by professionals using state standards
               Recognized as having poor social skills by their teachers
               Poor (below grade level) academic performance
6.     Identified a large number of participants (N>30?)  Six years to 22 years of age
7.     Collected Large data samples from each in authentic conversational dyads
8.     Transcribed the samples and then cycled through them listening and identifying
           behaviors that were problematic (seemed to interfere with the conversation) these
            would serve as potential functional indices of difficulty
9.     Isolated each of these behaviors:  labeled them and wrote a definition for each.
10.   Had 27 problematic behaviors at this time
11.   Created and applied a set of criteria to "refine" these behaviors
          -- Had to have occurred in at least two participants
          -- Had to have psychological reality (others could agree and easily ID them)
          -- Had to meet at least one of the following:
                * the behavior is a skill that should develop in normal communicators by 4 years
                    of age.
                * the behavior had been independently identified as a characteristic of
                    conversational difficulty by researchers
                * the behavior was found to have some discriminative power as demonstrated by
                    some experiments by Damico
12.     Reduced or refined the behaviors down to 17
13.     Searched for a theoretical Framework to employ and to anchor the descriptive
            behaviors  to an acceptable theory   (H.P. Grice's Cooperative Principle)
14.     Asked two groups (SLPs and Linguists) to conduct a "forced choice" and place the
             behaviors under one of Grice's four Conversational Maxims.
15.     Designed a procedure for "other behaviors" that might be identified
16.     Field tested the procedure and developed interpretive screens for rich interpretation
17.     Collected some data on occurrence of "temporal mapping problems" in typical
            communicators to determine "normal range of variation".
18.     First presented at ASHA as a miniseminar in 1980. Published in 1985.

                                                      The Cooperative Principle
                                                            (H.P. Grice, 1975)

                                            Categories of the Cooperative Principle

    1.      Quantity:   The quantity of information to be provided.
             A. Make your contribution as informative as is required
             B. Don't make the contribution more informative than is required
             C. Conciseness
    2.     Quality:     Try to make your contribution one that is true.
             A. Do not say what you believe to be false and represent it as true
             B. Do not say for which you lack adequate evidence
    3.     Relation: Be relevant.
             A. Your contribution to conversation should be appropriate to the immediate needs
                 at each stage of the transaction.
    4.     Manner: This is the act of relating not to what has been said but, rather, to HOW
            what is said is to be said.
             A. Avoid Obscurity of expression
             B. Avoid Ambiguity
             C. Be brief and orderly

                             CLINICAL DISCOURSE ANALYSIS:  Definitions

QUANTITY CATEGORY
        1. Informational Redundancy  ( RED).
             This involves the continued and inappropriate fixation on a proposition. The
             speaker will  continue to stress a point or relate a fact even when the listener has
            acknowledged its  reception and tried to proceed.
              Example    Examiner:  ...any way, I'm glad you enjoy the fair. Let's talk about
                                something else. How do/
                                "Did you ever see the bicentennial fair?
                                Examiner:  No, didn't see it.  Hey, do you like your teacher?
                                 "She's really OK. She lets me work on my bulletin
                                    board...she also lets me play with the cars."
                                Examiner:  The cars? Which cars are those?
                       "The model cars in the state fair exhibit. How much do you really
                                    like the state fair?"
        2. Failure to Provide Significant Information to the Listener  (FSI).
         The speaker does not provide the amount or type of information needed by the
          listener  for comprehension.
              Example    Examiner:  So, how would I get to your house from here?
                       "Turn right there where we play baseball and my house is down a
                                    little bit."
        3. The Use of Nonspecific Vocabulary  (NSV).
         The speaker uses deictic terms such as "this", "that", "then", "there", pronominals,
        proper nouns, and possessives when no antecedent or referent is available in the
        verbal or  nonverbal context. Consequently, the listener has no way of knowing what
        is being  referenced. Individuals displaying this difficulty also tend to overuse generic
        terms such  as "thing" and stuff" when more specific information is required.
              Example   Examiner:  Well then, what is your favorite toy?
                       "My favorite thing is...oh, stuff"
        4. Need for Repetition  (NR).
         Repetition is required prior to any indication of comprehension in spite of the fact
         that the material is not apparently difficult.
             Example    Examiner:   What did the little boy do then?
                                     "........................."
                                Examiner:   What did the boy do then?
                                     "...wh...What?"
                                Examiner:  When he saw this (points to picture) what did the boy do?
                           "he ran"

QUALITY CATEGORY
        5. Message Inaccuracy (MI).
         An attempted communication involves the relating of inaccurate info.

RELATION CATEGORY
        6.  Poor Topic Maintenance  (PTM).
         The speaker makes rapid and inappropriate changes in the topic without providing
         transitional cues to the listener.
              Example     "...but I missed it (an early T.V. program) cuz I went to bed."
                                Examiner:  That early? You must have had a hard day.
                       "Yeah"
                                Examiner:  What made it such a hard day?
                       "the raking"
                                Examiner:  That's hard work isn't it?
                       "Our teacher said, uh...whoever wins in checkers--I won!--goes
                                    to McDonald."
        7. Inappropriate Response (IR).
         The individual makes a response that indicates a radically unpredictable
            interpretation of  meaning. It is as though the individual were operating on an
            independent discourse agenda.
              Example Examiner:  How do you like school?
                       "I don't know him yet."
        8.  Failure to Ask Relevant Questions (FQ).
         The individual does not seek clarification of information that is unclear.
            Consequently,  there is little or no "verbal play" or clarification if the message
            received from the speaker  is unclear or too difficult for the individual to
            comprehend.
        9.  Situational Inappropriateness.(SI).
         The behavior tends to account for a generalized lack of relevance.  The speaker's
            utterance is not only irrelevant to the discourse, but it may also occur in an
            inappropriate social or interactional situation.
              Example Examiner:   Come in and sit down, Rich.  It's nice to meet you.
                       "Why does twenty years go by so fast?"
                            Examiner:  Pardon, what did you say?
                       "Why does twenty years go by so fast?
                            Examiner: .....Why do you ask that question?
                                 "I don't know...Do you always blink your eyes like that?"
        10.  Inappropriate Speech Style (ISS).
         The speaker does not change the structural, lexical, or prosodic form of his
            utterances  according to the needs of the audience or the context.  This may
            involve the occurrence of dialectal structural forms, code switching, style-
            shifting, language transfer, or interlanguage phenomena or idiosyncratic
            language codes

MANNER CATEGORY
        11.  Linguistic Nonfluency (LNF).
         The speaker's production is disrupted by repetitions,  unusual pauses, and hesitation
             phenomena.
                          Example    "sh...uh..she..um..she comes at dinner."
        12.  Revision Behaviors.(R).
         The speaker seems to come to dead ends in a maze, as if starting off in a certain
            direction, then coming back to a starting point and beginning anew after each
            attempt. There may be false starts and self-interruptions.
                          Example     "Well, you see...if you want-- sometimes when you ca--
                                            a lot of times when you can't go out, you can just play
                                           with your twin brothers."
        13. Delays Before Responding (DR).
         Communicative exchanges initiated by others are followed by pauses of inordinate
            length at turn-switching points.  This does include delays in responding to
            questions.
                          Example Examiner:   Well, what did you do at recess?
                                             "........played tag."
        14.  Failure to Structure Discourse (DS).
           This problematic behavior is the most global of the 17. It occurs when the discourse
            of  the speaker lacks forethought and organizational planning. Due to this
            characteristic, the  discourse is confusing--even if all of the propositional content
            is present.
        15.  Turn-Taking Difficulty (TTD).
         The participant in a conversational interaction does not attend to the cues necessary
            for  the appropriate exchange of conversational turns. This results in one of two
            possible outcomes. First, the individual does not allow others to add information.
            This is characterized by interruptions or consistent/inappropriate bids for a turn
                        Example   Examiner:  Well I think that the be/
                                   "I like the green one best."
                                        Examiner:  Yes, that's a nice one. How about the red one?" Do
                                            you wa/
                                  "Can we find more like this one?"
             The second possibility involves an opposite reaction. Rather than always bidding
                for a turn, this individual does not read the switching cues appropriately and
                therefore, does  not hold up his/her part of the interaction.
                      Example Examiner:   ...so this school year is almost over.
                                   "...it's not over though is it?"
                                    Examiner:   Not over, no it's not.
                                   "What do you mean by that?"
                                    Examiner:   Well, most of the year is gone...but not all of it.
                                            You know?
                                   "..................."
        16.  Gaze Inefficiency (GI).
             The individual's use of eye contact is inconsistent or uncharacteristic of gaze
                utilized in  the mainstream cultural context used as the criterion during this
                assessment.  Consequently, the attempts to "color" communication or help
                direct the interaction with  gaze are inappropriate or ineffective.
        17.  Inappropriate Intonational Contour (IC).
             The speaker's ability to embellish or "color" or contextualize his meaning
                through linguistic/tonal suprasegmentals such as pitch levels, vocal intensity
                and other inflectional contours is ineffective or inappropriate.

 GUIDELINES FOR USING CLINICAL DISCOURSE ANALYSIS
______________________________________________________________________________

ELICITATION AND DATA COLLECTION
1. It is essential that the data collection involve conversational interaction.   This
    procedure is designed to focus on the give-and-take of conversation and not picture
    description, narratives, question-and-answer formats or close-ended interviews.
2. It is important that the conversational partner (of the targeted individual) play by the
     actual "rules of conversation" and function as a true conversationalist. From a practical
     standpoint this might mean:
      --  both partners might provide a large amount of talk
      --  most topics will naturally arise and develop
      --  you don't violate "felicity conditions". For example, you don't feign ignorance of
            something that you know or you don't ask the targeted individual to perform
            redundant or meaningless tasks.
      --  Much of what is discussed comes from the "here and now" or from everyday or
            recent past experiences.
      -- Share as much information yourself as you expect the targeted individual to share.
3. When addressing the issue of "how much data to collect?", don't focus on the idea of
     obtaining a sufficient number of utterances. Rather, focus on AUTHENTICITY of the
     data. That is, how much confidence do you have that the data sampled really is an
     indication of the individual's typical performance and his/her best performance? There
     are some guidelines that can typically be used to increase the potential of authentic
    data:
      --  Collect data from a full range of speech events
      --  Collect data from recurrent instances of those events
      --  Collect data using several data collection approaches.
     Never rely on Clinical Discourse Analysis by itself to provide you with the data
     needed to describe the meaning-making proficiency of an individual.
 4. Authenticity is not determined by a number of utterances or contexts, it is determined by
     your degree of judgment and confidence is the representativeness of the behavior. It is
     best to collect conversational data from several instances of conversation. The fewer
     samples and contexts sampled, the greater the chance of non-authenticity.
5. Don't think of context as only involving physical setting. Contextual variables also
     include
      --  social setting
      --  overt behaviors used to frame and organize
      --  language phenomena spoken by both parties
      --  extra situational variables including background knowledge, experiences and
            motivations.
6. The context is very dynamic and frequently changes instanteously.
7. The conversational interaction should be tape-recorded so that it can be transcribed and
     analyzed at a later time.

SEGMENTATION OF THE SAMPLE
1.  The segmentation is accomplished using two types of data from the sample:
  A.  Phonological Unit - a segment of speech separated by the contours of intonation,
          stress, and a pause in the subject's voice. This is usually marked by a definite pause
          preceded by a diminishing of force and a drop in pitch of voice (or a rise for
         queries).
 B.  Communication Unit - a segment of speech containing and defined by the   presence
        of a proposition that provides conceptual information (as expressed by a predicate
        -- explicit or implied -- taking one or more argument). Typically, this unit can't be
            broken down or further divided without loss of essential meaning   (but see note
            2.2 below).
2. The phonological and communication units usually co-occur. When this is not the case,
         always segment according to the meaning component (communication unit).
          #Meaning#  (terminal juncture, meaning unit, tj) = 1 unit
          #Meaning, Meaning# = 2 units of transcription
          #partial meaning# Meaning# = 1 unit of transcription
          #partial meaning, partial meaning, Meaning# = 1 unit
 2.1- Communication Units need not be complete structurally
          "Where are you going?"         "Home"   (1 unit)
 2.2-  Compound sentences that contain only 1 conjoining device and that fall within a
          phonological unit are counted as one utterance (e.g., "and", "then", "and then").
                 "I'm going to the store and I will buy bread"
 2.3-  Conjoined sentences that have undergone subject or predicate deletion are counted
        as one utterance.
                  "I was coming home but ran into a big snowstorm"
 2.4- Adjoined utterances are counted as one utterance provided they are within one
          phonological unit.
           --Temporal link: before, after, when, while
           "After its put in cans, its shipped out"
           --Casual link: because
           "She let me stay because I was broke"
           --Conditional link:  if
           "If its cold we might have a frost"
           --Purposive link:  so
           "The oil repels water so it can swim"
           --Disjunctive link:  instead, although, but
           "Instead of getting some sleep, they went fishing"
           --Parenthicals and asides are counted separately
           "I have a pretty star (it blinks!) on my tree"

TRANSCRIPTION OF THE SAMPLE
1. Preserve the speaker-hearer dyad. This means collecting and transcribing behavior
         from both interactants.  It is helpful to do so in a regular format.
2. Retain all non-linguistic contextual information.
3. Preserve the contextualization cues
4. Utilize the key on the form as a standard coding procedure.
5. Number the line in the transcription.
6. Place different transcription units on different lines.
7. Keep the transcription accurate across all interactants.
8. Do not omit any "verbal segments" that seem incomplete.  These false starts and
        mazes  are important indices of temporal mapping problems.

CODING OF THE BEHAVIORS IN THE SAMPLE
1. It is essential that all behaviors are coded. Do not engage in the process of
        interpretation at this stage.
2. Use the standard codes for the problematic behaviors (See form).
3. Use the Clinical Discourse Analysis definitions to  identify and code the problematic
         behaviors.
4. Place the analysis on the actual transcript. The standard symbology (see form) should
         be  used and the behavior should be coded by placing the symbol over the behavior. 5. Remember that description and identification of problematic difficulties is the primary
         object of the analysis. If a behavior can be validly described by the use of two
         symbols,  then use them.  It results in more subtle description.
6. If there is a problematic behavior that cannot be coded with one of the CDA
         behaviors, code and document it using the following strategy:
  A.  Identify the behavior and describe it in writing.
  B.  Provide a descriptive term that uniquely identifies it.
  C.  Determine which Gricean category it best fits.
  D.  Add the problematic behavior and code it with the others
7. After initial coding, it is frequently helpful to listen to the tape again while treading the
         transcript. Check your transcription and your coding -- particularly the temporal
        mapping  problems.
8. Beware of the tendency for "prescriptionism" in coding.  You must code what is present,
        but don't be overly judgmental.
9. There are a number of "pitfalls" in the behavioral coding. Some information to assist in
        accurate coding is listed below by problematic behaviors:

 Informational Redundancy
 -- Key: "Continued and inappropriate fixation".
 -- Usually revealed as a limited number of topics or a limited repertoire of topics.
 -- Grammatical redundancy is not informational redundancy.
 -- Don't confuse a repeat for clarification or emphasis or re-establishment of a previous
            topic as this behavior.
 -- A repeat is not a fixation.

 Failure to Provide Significant Information
 -- Remember to focus on content not form. For example, a complete sentence is not
      necessary for meaning transmission. Ellipsis is not a problematic behavior. Don't
      code it as one.
 -- This behavior is most easily observed when asking for specific information or when
        the individual spontaneously provides directions or instructions.
 -- Note antecedents or linguistic context to provide information. This is acceptable.
 -- Key:  If you know what the child is talking about...even if the information doesn't
          appear to be available, then it is not a problematic behavior.

 Use of Nonspecific Vocabulary
 -- Generic terms are acceptable for glossing over.
 -- Generic terms are acceptable when utilizing nonverbal or other contextual information.
 -- The key phrase in the definition is "when no antecedent or referent is available"
 -- Make certain that you only code initial instances of nonspecific pronominals.
 -- Key:  If you know what the child is talking about...even if the information doesn't
        seem to be available, then it is not a problematic behavior.

 Need for Repetition
 -- When you have asked a question  and  you  must rephrase, restate, or move to
        another question or topic, this is typically a manifestation of this behavior.

 Message Inaccuracy
 -- Key: Inaccuracy or mis-information
 -- The question has been addressed but the response (though a potential response) is
        inaccurate
 -- May note this as an obvious contradiction of the context
 -- It doesn't matter if the person believes that he is relaying false information or not......
        the key is inaccuracy.
 -- Incorrect answers in class are coded here.
 -- Not a gauge of moral integrity.
 -- Confabulation within the same question frame is coded here

 Poor Topic Maintenance
 -- This can also be due to a poor transition as well as no transition....perhaps its triggered
        by a somewhat related idea.
 -- Typically find that the topics cannot be sustained -- even when attempted
 -- Distractibility may be one "general feeling" when this is a problem
 -- If there is a feeling of "losing your verbal feet", note whether this is due to a rapid or
          inappropriate topic shift
  -- Even if a reason for the shift is obvious, if it was a poor transition, code it now and
          interpret later.

 Inappropriate Response
 -- Key: Focus is on a plausible response
 -- Inaccurate responses that are within the same question frame are not coded here.  See
          message inaccuracy
 -- The key phrase is "radically unpredictable interpretation"
 -- if you must reassert the propositional Question or it is lost -- code it here
 -- When this relational problem comes after a query, code it here. This is a way to
            distinguish between PTM, SI and IR.

 Failure to ask relevant Questions
 -- This is based more on lack of behavior
 -- More interpretive
 -- Ask the question:  Does the individual try to obtain clarification or make repair when
        he doesn't catch the question or doesn't understand the verbal environment?
 -- Types of behaviors linked to this lack of clarification would be:
           no response
           inappropriate response after a delay
           repetition of Q or of his previous utterance
           stereotypic responses
           non-attentiveness

Situational Inappropriateness
 -- A generalized lack of relevance
 -- A more all-encompassing category
 -- due to social or interactional inappropriateness
 -- may be reflected in distinct social "faux pas" or extreme boorishness
 -- often does not read social contextualization cues
 -- violations of social constraints
 -- A more "catch-all" category

 Inappropriate Speech Style
 -- This is the category for coding difficulties in the superficial aspects of
        language structure
 -- Code switching that is inappropriate is coded here.
 --  Language transfer phenomena and dialectal differences are coded under this category.
 -- Idiosyncratic terms and language structure coded here.
 -- Inappropriate speech registers and style shifting are included here.

 Linguistic Nonfluency
 -- Major caution: Count all behaviors. Use a "trips the ear" criteria.
 -- Beware of multiple coding
 -- temporally contiguous instances are counted as one instance. Particularly if they
        haven't moved past some propositional information.
 -- Note that there are filled LNFs.
 -- Review the tape after transcription to help find these
 -- We all produce these behaviors. It appears that the frequency of occurrence is the
        key to difficulty. You must code all instances and then interpret.

 Revisions
 -- "Mazing" is a key descriptor
 -- Even phoneme revisions are noted and coded.

 Delays before Responding
 -- Can have a filled DR
 -- Simply code the occurrences.  Issues of  thoughtful pauses, multiple questions,
        complex syntax and processing thought-provoking queries will be handled in the
        interpretation phase.
 -- Nonfluency before answering a question is a DR

 Structuring Discourse
 -- Usually suprasentential

 Turn Taking Difficulty
 -- there can be two types:
       Violating constraints and cues to inappropriately take the turn
       Not upholding the turn
 -- Violations of Adjacency principles coded here
 -- Code all turn breaks.
 -- Note if an individual continues on inappropriately when overlap occurs.
 -- Don't place back-channels or assessment reactions here.

 Gaze Inefficiency
 -- This is very culturally determined.
 -- Focus on what the context requires and then interpret after you code according to
        the context.

 Inappropriate Intonational Contour
 -- This does include intonational performance due to motoric problems.
 -- Flat or inconsistent contour typical occurrences
 -- Key is when the result is mixed messages or inefficient contextualizing

INTERPRETATION OF THE DATA
 Key Concepts:
      Describing Authentic Social Action
      Pragmatic Mapping
      Forced Adaptations
      Patterns and Systematicity
      Subtle Complexity

1. This phase is HEAVY on your knowledge and informed judgment.  Don't be
    intimidated.   Exposure and experience will make you better.
          ** Read the literature
          ** Analyze some normals
          ** Gain experience with procedure
          ** Embrace the complexity
2. Once you gain more experience, the screens an be employed quickly and effectively.
3. There must be a framework built into the assessment process that allows the SLP to
    account for  conversational complexity.
4. It is not enough to recognize the fact that problematic performance may be due to
    differences or strategic reasons as well as deficits....... you must be able to
    systematically account for this  possibility in a way that is "institutionalized".
5. There is a tendency for prescriptionism in interpretation as well....use the
    questions/screens  listed below to help you prevent this tendency from influencing
    your interpretations.
6. Don't be too concerned if you miss behaviors.  The pattens will reveal themselves if
    you are  systematic and if problems actually do exist.
7. You may use both qualitative (discussion of patterns and behaviors) and quantitative
    (numbers  and percentages) data.  Don't over-extend their interpretive power.
      **  numbers alone can't determine level of difficulty or whether or not there is a
               problem or disability.
      ** the patterns and types of behaviors are more important
      ** the changes in occurrences of behaviors may be important
8. It is key to remember that the behaviors on Clinical Discourse Analysis do serve as
    effective  indices for potential problems during conversation.  They may serve as "road
    maps" telling you  where to look for problems.  They may be viewed as potential
    "Forced Adaptations".  But this is an interpretation accomplished with completion of
     the bi-level analysis paradigm.

         Two Distinct Levels of Analysis
          A. Descriptive Analysis Phase
                Focus on Overt Behaviors that may serve as Indices of Problems
                Functional Difficulties and Strengths and Adjustments
                Clinical Discourse Analysis is one such procedure
                Identify and Code all Problematic Behaviors
                    Look for Patterns
                         1)  Collect all the data
                         2) Move through data and note co-occurrences
                         3)  Determine numbers and percentages
                         4)  Determine whether there is enough to consider potential problems
                               * Do these behaviors interfere with social activities
                               * Do they bother others
                               * Are there specific and troubling patterns
                               * Are there sufficient occurrences
                               * Are the types of behaviors unusual
                               * Do they prevent success of the three criteria
                                        Effectiveness
                                        Appropriateness
                                        Fluency
                         5) If problems exist from the perspective of the context of  interest,
                                then move to the second analysis level.

          B. Explanatory Analysis Phase
                Focus on Causes and ask Why do the problematic behaviors occur?
                This process typically involves the combining of LOW INFERENCE data
                    (observational data, test results, interview information on background, etc)
                    to form HIGH INFERENCE JUDGMENTS.
                There are sets of questions and considerations that can assist in addressing this
                process . This process will result in the following:
                      * A summary of the data collected
                      * A deeper interpretation of the data to comment on
                              1) whether there are difficulties
                              2) what areas of functioning are affected
                              3) whether these difficulties are due to impairment
                              4) what contextual variables are significant
                              5) what is the student's intervention potential
                Process
                     1. Take each of the problematic behaviors (individually or the pattern)
                     2. Analyze it (them) to see why they occur
                     3. Start with an assumption of normalcy
                     4. Apply the questions of screens detailed below

9. The following Explanatory Questions/Screens are designed to force you to consider
    the potential causal factors for the occurrence of the problematic behaviors in the
    conversational  sample.  These are designed into several general categories of
    consideration.  Proceed in the  following order.  This will prevent you from mistakes
    due to a prescriptionistic bias.

     Screen One:  Are the problematic behaviors manifestations of Normal
                            Conversational Strategies?
        -- Interactional strategies are often used to "opt" out or to maintain control.
             Ambiguity (FSI....RED...NR)
        -- Defense of Face
        -- Discourse markers to gain control (LNF....R)
        -- Avoidance manifested as Poor Topic Maintenance
        -- Word Search as attention getting device
        -- Forgetting as an interactional resource
        -- Nonfluency to maintain the floor
        -- Mid-gaze or "Thinking Face" to signal need for co-participation
        -- Message inaccuracy as bantering or to get a reaction

 Screen Two:  Are the problematic behaviors merely in random variation or
                        operating within normal limits?
        -- Temporal Mapping Problems are especially relevant here (Linguistic Nonfluency,
                Revisions, Delays before Responding)
                ** Normal range of occurrence (6% - 24% of utterances)
                ** Can increase these in all of us by operating outside of  the level of
                        comprehensibility and/or comfort
                ** Best to note significant change in occurrence with Concomitant change in
                        level of complexity
        -- There does not seem to be a pattern or repeated occurrences
                ** Simple communicative breakdown
        -- Delays before responding as a reaction to multiple questions, thought-provoking
                queries, complex syntax and processing.

Screen Three: Are the problematic behaviors due to procedural mistakes by the
                        assessor?
         -- Incorrect Coding Procedures
              ** Too Prescriptionistic or biased
              ** Multiple NSV on pro-nominals
              ** FSI for lack of lexical specificity despite comprehension
              ** Message inaccuracies as Inappropriate Responses
        -- Violation of Sincerity Constraints
        -- Operated beyond the level of comprehensibility
              ** Give Rise to Temporal Mapping Problems
        -- Not a conversational interaction
        -- Inappropriate contextualization cues

 Screen Four:  Are the problematic behaviors due to language or dialectal differences
                        between the individual and the language code of the context of
                        interest?
        -- Inappropriate speech style due to VBE or code switching
        -- Culture of Poverty and "restricted code"
        -- Note the problems in L2 but not in L1
        -- Due to interlanguage phenomena
        -- Due to normal L2 acquisition phenomena
        -- Due to language transfer errors

 Screen Five:  Are the problematic behaviors due to cultural differences/
                    interferences that affect interactional expectancies or strategies?
        -- Different eye gaze
        -- Different nonverbal contextualization
        -- Different clarification strategies
        -- Silence as a conversational device

 Screen Six:  Is there an indication of extreme test anxiety during the observational
                assessment in one context but not in others?

 Screen Seven: Is there significant performance inconsistency between different
                observation periods in the same context?

 Screen Eight:  Have any contextual variables of significant been identified?

 Screen Nine:  Is there any evidence that the problematic behaviors noted can be
                explained according to any bias effect that was in operation before,
                during, or after the assessment?
        -- Is the student in a subtractive bilingual/cultural environment?
        -- Is the student a member of a disempowered community?
        -- Are negative or lowered expectations for this student held by the student, the
                student's family, or the educational staff?
        -- Were specific indications of bias evident in the prereferral, referral,
                administrative, scoring, or interpretative phases of the evaluation?

 Screen Ten:  Are the Problematic Behaviors Potential Compensatory Strategies?
        -- "Isy" as a discourse marker
        -- Simmons-Mackie & Damico,  1996; 1997

 Screen Eleven: Are there any remaining (unaccounted for) Problematic Behaviors
            or Patterns of behaviors that reveal an underlying linguistic systematicity
            during the descriptive analysis phase?
        -- This question applies only to the problematic behaviors that are still remaining
                after application of the ten screens.
        -- Isolate turns/utterances containing remaining problematic behaviors
        -- Is there significant performance inconsistency between different input or output
                modalities?
        -- Perform a systematic linguistic analysis on these data points looking for
                 consistency in appearance of problematic behaviors.
              * Grammatical (primarily syntagmatic)
                     Crystal profiles (1982)
                     SALT (Miller & Chapman, 1983)
              *  Semantic (primarily paradigmatic)
                     PRISM (Crystal, 1982)
                     Perceptual/Language Distancing
                     Displacement
                     Levels of Abstraction
                     Cohesion/Coherence (Halliday & Hasan, 1976)
        -- Look for large changes in the occurrence of the problematic behaviors as the
            complexity of one of these dimensions increases (e.g., an increase in
            grammatical complexity from LARSP Level IV to LARSP Level V causes an
            increase in the number of linguistic nonfluencies from 18% to 39%.
10. Determine Appropriate Placement
     * A plan of action detailing what further service delivery should occur.  This should
        include:
         1. What should be done
         2. Who should do it
         3. When services are provided
         4. What support will be provided
         5. How follow-up is accomplished