Grasshopper Enterprises

xperimental division

Dr. Charles P. Reinert’s Study on Image-Streaming and IQ


A special one quarter, 4 credit hour course was developed at Southwest State University in order to begin to understand the effect of a verbally described imagery process, generically referred to as Image Streaming, on the development of intellectual skills of university students. Most of the students in the course had been provisionally admitted to the University, with a measured I.Q. slightly below 95. Pretests and posttests of analytical skills, creativity, and learning style were administered. Students’ verbalization techniques were monitored during each class. Cerebral dominance was measured using eye, ear, and leg preference. Occasional feedback was solicited from students concerning health, the number of intuitive insights experienced and other factors.

Preliminary analysis of results suggests that students’ analytical skills rose with increasing hours in image streaming, with the largest rates being measured for the lowest initial analytical skills. The corresponding I.Q. gain per hour of practice ranged from a high of +2.3 I.Q. points per hour to a low of -0.9 I.Q. points per hour, with a standard deviation of 0.7. There was some indication that students with the highest I.Q. gain rates tended to be left cerebral dominant, those with intermediate gain rates were mixed dominant, and those with the lowest rates were right cerebral dominant. The average gain was 0.44. Increases in “artistic” creativity were also noted, with slightly larger increases noted for students with initially higher analytical skills. These students also reported more intuitive insights than the students with initially lower analytical skills. A modest decrease in “verbal” creativity was noted, this decrease being slightly greater for those students with initially higher analytical skills. As a whole, the group moved slightly toward preferences for “active experimentation” and “concrete experience”, as measured by Kolb’s Learning Style Inventory. Limitations of the study are discussed.


In the winter 1988-89, this author undertook a first preliminary study of the effect of “Image Streaming” upon the performance of students in a general education level physics course at Southwest State University in Marshall, MN. (“Image Streaming” is a term coined by Dr. Win Wenger, president of the Institute of Visual Thinking of Gaithersburg, MD., who developed and refined the image streaming process. Strictly speaking, image streaming applies to only the imagery//verbal description process associated when no “trigger” is used, as described later.) In this first study, students with an average I.Q. of 106 were given initial instructions for image streaming, checked twice thereafter, but otherwise did all of their image streaming out of class, on their own time, and kept their own time records.

The results of the first study, suggested a positive correlation between hours of image streaming and an increase in students’ analytical skills, as measured by a simple 38 point test, the Whimbey Skills Inventory. This “WASI” test had previously been correlated with the Otis Lennon Mental Ability Test (A. Whimbey, private communication). On the basis of the correlation, the resulting I.Q. increase was found to be approximately 0.8 I.Q. point per hour of practice. It was also found that the average learning style of the students who image streamed moved toward a more “balanced” position, as measured by Kolbs Learning Style Inventory (Kolb, 1976). Later analysis of this data indicated that the I.Q. gain rate dropped somewhat with increasing initial I.Q.

In the fall of 1989, this author again attempted to measure the effect of image streaming, this time dedicating an entire 40 clock hour university course to the process. The course, remedial in nature, met for one hour each day, 4 days per week, in the same, comfortable room as used for one of the sections in the earlier study. The routine was approximately as follows: Relaxing music (“Crystal Suite” by Steven Halpern or similar) was used during each class, played by CD system through a stereophonic sound system, and at a level sufficiently low that students could easily converse above it. At the beginning.of the class, 5 minutes of simple stretching exercises were used to prepare the students for class. This was followed by a 20 minute period during which the students “image streamed”: The process was one of typically closing the eyes, then describing to a partner (each in turn), the images which appeared “before the eyes”. Specific instructions were given each class period that the description was to be very detailed, and that students were to attempt to describe using all five senses, and in the present tense. In approximately 1/3 of the class periods, a “trigger” was used to encourage the students to “get started”-- for example, the students were invited to describe a “beautiful garden” as the first imagery exercise. In a later session, they were invited to image receiving a “letter from NASA”, with an invitation to participate in a voyage to Mars, etc. Students had the option of using the trigger or not; most did when it was suggested. Following the twenty minute period, students were asked to rate the experience on a scale of 1 (worst) to 10 (best). They were then to spend the next fifteen minutes writing about their imagery and and then fifteen more minutes sketching (with colored markers) their imagery. Attendance at the class sessions averaged approximately 75%. As “homework”, the students were to originate two more imagery sessions, done in the same way, with or without a trigger as they chose. Students kept track of their own time spent in the imagery process.

Occasionally during the ten week class, student feedback was solicited concerning physical/emotional health, attitude toward the class, and number of intuitive “insights” experienced recently.


A. Average Results:

  1. Number of students assessed: 24
  2. Average time in image streaming: 20.5 hours
  3. Average attendance, percent: 70%
  4. Average entry level I.Q.: 94
  5. Average exit level I.Q.: 103
  6. Average I.Q. gain rate: 0.44 IQ pt/hour
  7. Entry level creativity scares:
    Guilford “Decorations” 38.9
    Guilford Expressional Fluency: 4.2
  8. Exit level creativity scores:
    Guilford “Decorations” 49.6
    Guilford Expressional Fluency: 3.6
  9. Entry level Kolb coordinates:
    Active Experience - Reflective Observation: -1.2
    Abstract Conceptualization - Concrete Experience: 0.8
  10. Exit level Kolb coordinates:
    Active Experience - Reflective Observation: 1.0
    Abstract Conceptualization - Concrete Experience: -0.2
    B. Analysis Of Trends

In an effort to sift the data for trends, the results were divided into two groups, based upon their rate of Increase of I.Q. with time. The 13 students with the highest rate of Increase with time (I.Q. gain rate) are identified as the “high 13”; those 11 students with the lowest rate of gain are identified as the “low 11” in the results following:

High 13 Low 11

  1. Entry level I.Q.: <90 101
  2. Cerebral Dominance: 5L, 5M, 2R 5L, 2M, 2R
  3. Gender distribution: 7M, 6F 8M, 3F
  4. Entry level “Decorations” 35.6 42.8
    Entry level “Expressional Fluency” 4.1 4.4
    Entry level Kolb L.S.I.: AE-RO: -2.8 0.9
    Entry level Kolb L.S.I.: AC-CE: 2.3 -1.1
  5. Exit level “Decorations” 54.8 45.2
    Exit level “Expressional Fluency” 3.6 3.6
    Exit level Kolb L.S.I.: AE-RO: 1.0 0.9
    Exit level Kolb L.S.I.: AC-CE: 1.6 -2.3
  6. Average I.Q. gain rate 0.7 0.2
  7. I.Q. gain rate by linear regression No correlation 0.9 pts/hour

The average rate of Increase in I.Q. as measured by the 38 point Whimbey Analytical Skills Inventory was found to be 0.44 IQ points per hour of image streaming practice. What may be more indicative, however, is a mathematical fit to the data. The functional fits of the IQ change versus hours of practice via linear regression analysis are as follows, for the two subgroups.

1.“Lower 11”:

WASI change (WC) versus I.S. hours (ISH):
WC = -12.3 + 0.62 ISH
Coefficient of determination = 0.25
Coefficient of correlation = 0.50
Standard deviation = 11.1

  1. “Upper 13”: WC = 27.3 – 0.41 ISH

Coefficient of determination = 0.117
Coefficient of correlation = -0.34
Standard deviation = 9.86.

The correlation for the “lower 11” is considered to be sufficiently, high that one can place some trust in the fit. In this case and in view of the 1.5 ratio between IQ change and WASI change, the IQ gain rate becomes slightly over 0.9 IQ points per hour of practice. Note that the coefficient of determination is not large. (A COD of 1.0 would be “perfect”.) The mathematical slope of the function is 0.54, comparable to the value of 0.62 obtained with the 1988 study. The large value of the “constant”, -12.3 in the mathematical fit for the “lower 11” suggests that, in this case, about 13 hours of image streaming were required before any IQ gain began to show.

The correlation for the “higher 13” is seen to be negative, though of a lesser magnitude. and therefore less reliable. Note that the coefficient of determination in this case is only 0.117, and therefore the mathematical function cannot be considered very reliable.


  1. Change in “Decorations” (DECC) with image streaming hours (ISH)
    For “lower 11”:
    DECC = 0.18 + 0.71 ISH
    COD = 0.1
    COC = 0.32
    SD = 16.4
    For “upper 13”:
    DECC= 9.02 - 0.03 ISH
    COD = 7.7 x 10-4
    COC = 0.03
    SD = 6.7

  2. Change in “Expressional Fluency” (EFC) with image streaming hours (ISH)
    For “lower 11”:

EFC = 0.25 -0.023 ISH
COD = 0.01
COC = -0.11
SD = 1.3
For “upper 13”:
EFC = 0.75 -0.055 ISH
COD = 0.014
COC = -0.12
SD = 2.8

Evidentially, there is a modest correlation between image streaming hours and the “Decorations” score for the “lower 11” group, but a negligible correlation for the “upper 13” group. The correlation is negligible for both groups with the “Expressional Fluency” test.


At this stage of analysis, a model which fits all of the data has not suggested itself to this author. Simplistically speaking, however, I suggest the following for consideration:

A. For students with IQ’s above 100 (and perhaps the absence of clearly defined “learning difficulties”), there seems to be a reasonable, positive correlation between IQ gain as measured by the 38 point Whimbey Analytical Skills Inventory and the hours recorded by students as spent in image streaming. The rate of gain is in the vicinity of 0.9 IQ points per hour of image streaming practice, which is consistent with (even somewhat greater than!) the rate of gain measured in the author’s earlier work. (One should bear in mind that students normally spent some additional time in writing about, and in drawing, their Images following the image streaming exercise per se.)

B. For students with IQ’s below 100 (and perhaps additionally with "learning difficulties), there appears to be much more scatter in the data, though the larger IQ gains do appear in this group. The mathematical slope of the the “best fit” line is actually negative for this group, however the goodness of fit is much poorer than the the other group. Perhaps there was a good deal of experimentation, reorganizing, what have you, happening for these students.

C. There is a modest, positive correlation between creativity as measured by the artistically oriented “Decorations” test and image streaming hours for the “lower 11” group. This suggests that image streaming has a positive effect on some types of creativity for some IQ groups. On the other hand, the “upper 13” group (lowest entry level IQ’s) had no such correlation. Apparently, creativity gains for lower IQ’s do not change rapidly with image streaming practice. Following the suggestion of Win Wenger (personal communication), perhaps "what needs fixing worst gets fixed first’'-- It may be that IQ is the first quantity to change, and when this has increased sufficiently, positive changes in creativity begin to occur.

D. Clearly, much more work needs to be done. This author is presently compiling additional data from other classes where image streaming was used, and colleague Win Wenger has a major study in progress as well (personal communication). The limitations of this study are clear, at least to the author: IQ testing has been rudimentary and certainly not “standard”-- the 38 point WASI is convenient but does not have high status in the field. Also, the creativity tests which were used are “old”, and there may be much better ones now available. Attendance data for this work was inadequate, as was the method of allowing students to monitor their own time investment. Additionally, the conditions under which the class was conducted are not typical-- music and a generally low stress environment are, regrettably, not yet the classroom norm. (It is worth noting in this regard, however, that while student technique was closely monitored in this study, the image streaming in the earlier study was done entirely independently. Yet we achieved similar results in terms of the IQ gain per hour investment.) Finally, larger student numbers, and better data on student entry capabilities are in order.

From a personal perspective of working with approximately 200 students over 1 1/2 years, the author remains very impressed with not only the quantitative improvement, which seems to accompany the image streaming process, but also its ease of use. I have yet to work with a student who, when using proper technique, was unable to “get pictures”. Some are of course much better at the process than others, but it seems possible, and relatively easy, for all to successfully use this technique. Considering that, once the student has been taught the proper technique, no instructor seems really necessary thereafter, it is tempting to suggest that this technique may be a very useful one for assisting large numbers of students (e.g. thousands) in basic skills development. A basic 5 clock hour course in image streaming technique would seem more than sufficient to allow the motivated university freshman to continue skill his/her own skill development, perhaps to much higher levels than we are accustomed to thinking about for our students. Finally, the author is personally convinced that creativity increases do accompany the image streaming process, If for no other reason than from the accounts by surprised students of the intuitive insights which begin to occur after about the first 4 weeks of image streaming practice. Though difficult to measure, I’m convinced they are there. Image streaming may therefore be very useful in the inventive/problem solving process which we must value highly in this technological society. In view of at least suggestions that the gain may be larger for lower values, its use by the mentally impaired is also important to consider.