The Learning Behaviors Scale

1
The Learning Behaviors Scale

• P. A. McDermott, L. F. Green, J. M.
  Francis, D. H. Stott

2
Description of the LBS I

• 29 items, each presenting a specific
  learning-related behavior
• Observer is required to indicate whether behavior
  Most often applies, Sometimes applies, or Does
  not apply.
• Some items indicate positive learning behaviors
  and others indicate negative behaviors to reduce
  response sets.

3
LBS Description II

• The 29 items provide 4 subscale scores
• Competence Motivation (8 items)
• Attitude Toward Learning (9 items)
• Attention/Persistence (7 items)
• Strategy/Flexibility (7 items)
• Subscales allow for targeted intervention
• You can also obtain a global LBS score

4
CONTEXT FOR DEVELOPMENT I

• Knowing that a student is bright or not as
  bright does not fully explain performance
• Knowing a students intellectual capacity
  provides limited information for intervention
• What do you do when faced with a bright student
  who is not doing well?

5
Context for Development II

• There are behaviors associated with learning.
  What behaviors do you think are associated with
  learning?
• Some of them are as follows
• Listening attentively
• Participating in classroom activities
• Accepting correction
• Sticking to tasks until completed
• Working to please teacher

6
Context for Development III

• There is evidence that these learning behaviors
  are teachable.
• Idiographic data is useful in this arena, but
  time-consuming to gather
• We did not have good nomothetic data on learning
  behaviors.
• However, we know that school teachers are
  relatively accurate, reliable, unobtrusive,
  cost-beneficial observers of classroom behavior,
  when they have had ample opportunity to observe.

7
Context For Development IV

• In light of the previous information, the authors
  of the LBS set out to develop a scale to measure
  learning behaviors reliably and validly in 5 - 17
  year olds, using teacher observation.
• The research work on the LBS started in the mid
  1980s and the scale was published in 1999.

8
Preliminary Research

• Examined
• Reliability of subscale scores and total score in
  small samples.
• Validity of subscale and total scores in small
  samples.
• Created scale that worked well.

9
US Standardization

• Sample
• 1,500 students, 750 males and 750 females 5 to 17
  years old
• Used 1992 U.S. Census to obtain demographics
• Blocking for sex, age, grade in school
• Stratified random sampling by race, class, family
  structure, community size, and geographic region

10
Final Norm Sample

• Race 67.7 White, 15.9 Hispanic, 12.1 African
  American, 4.3 other groups.
• Family 76.6 with two parents or guardians,
  21.3 single mom, 2.1 single dad.
• Representation by SES based on parent education,
  and exceptionality
• Final selection was randomly selected from those
  who gave consent, restricted only by
  stratification quotas and 2 students per
  teacher.

11
Test Retest Reliability (n 77)

• Competence Motivation .92
• Attitude Toward Learning .91
• Attention/Persistence .92
• Strategy/Flexibility .93

12
Inter Rater Reliability (n 72)

• Competence Motivation .83
• Attitude Toward Learning .83
• Attention/Persistence .83
• Strategy/Flexibility .83

13
Internal Consistency I
14
Internal Consistency II
15
Validity Coefficients
16
T T Sample

• The LBS was completed on all 700 students in the
  sample, with no rater missing more than 2 items.
• The distribution was skewed toward the higher
  end--that is, most students were rated as having
  learning behaviors in the normal range.

17
Factor Analyses

• As is recommended, we used factor analysis to
  look at the structural validity of the LBS in the
  T T sample.
• We used multiple criteria to determine how many
  factors would work best.
• The goal was to find a factor structure that was
  generalizable across the whole sample, as well as
  the gender subgroups.

18
Factor Analyses 2

• In the US, the LBS is made up of four factors.
• We ran five factor, four-factor, three-factor,
  and two-factor models.
• The only structure that generalized from the
  whole sample across gender groups was the
  two-factor one (see p. 8 in manual and note
  pattern coefficients).

19
Factor Analyses 3

• Factor I consists of 18 items and is labeled
  Attitude Toward Learning (AL). The items on this
  factor made up the Competence Motivation,
  Attention-Persistence, and Attitude Toward
  Learning subscales in the U.S. structure.
• Factor II, labeled Strategy Flexibility,
  consisted of the 7 SF items in the U.S. norming,
  but of 9 items in Trinidad.
• The AL and SF factors have two items in common
  (10 14).

20
(No Transcript)
21
Attitude Toward Learning Items
22
Strategy Flexibility
23
Lets look at reliability estimates for the T T
scores
24
Reliability and Validity Evidence

• Reliability estimates for scores on the total
  scale were consistently high across all
  subgroups.
• Reliability estimates for subscale scores were
  very high for Factor 1 and moderate for Factor
  II.
• No reliability estimates fell below .75.
• There were no statistically significant
  differences between genders, among ethnic groups,
  or among grade levels.

25
Administration

• Speak to teacher who has seen students for at
  least 6 school weeks or 30 days.
• Ask teacher to rate the student as accurately as
  possible.
• Let the teacher know that this information will
  help in your assessment of the student.
• Teacher should rate all responses.
• Will require 5 to 10 minutes to complete.

26
Scoring

• Use scoring template to complete raw score for
  each dimension.
• Raw scores for Factor 1 (AL) range from 0 to 36.
• Raw scores for Factor II (SF) range from 0 to 18.
• Raw scores on Total Scale range from 0 to 54.
• Put raw scores in boxes on Score Summary sheet.
• Convert raw scores to tiles using table on p.
  10.
• Always double check your scoring.

27
Interpretation I

• Scores on the LBS should be only one part of a
  broader psychoeducational evaluation.
• Higher scores represent the presence of more
  learning behaviors.
• Students who obtain scores at or above the 40th
  percentile are displaying learning behaviors at
  or above the average range.

28
Interpretation II

• Students whose learning behaviors are between the
  20th and 40th percentiles are students who may
  benefit from interventions aimed at increasing
  their learning behaviors in general.
• Students whose scores fall below the 20th
  percentile are manifesting deficits in learning
  behaviors and may benefit from immediate
  interventions.

29
Lets Practice