Preface

This Guide for the Investigator was written separately from the FACS Manual for three reasons:

(1) We have assumed that anyone using FACS will want to have at least one other person learn the system so that inter- coder reliability can be determined. Since the other person might be someone not trained in scientific methods, we have kept out of the Manual, but placed in this Guide, any technical information requiring sophistication on the part of the reader.

(2) It would be ideal if the person coding facial actions did not think about the possible meaning of the behavior he scores, but focused only on describing appearance. Information about the meaning of facial behavior has been kept out of the Manual and is mentioned only in this Guide.

(3) FACS is an unavoidably expensive package, entailing not only a text but many photographs, motion picture films, computer programs, etc. It seemed reasonable to provide the prospective user an opportunity to find out something about FACS at a minimal cost to determine whether it is appropriate to his interests. The first five chapters of this Guide provide the prospective user with that information.

FACS is in the experimental stage of development. It is not complete, final nor proven. It is the product of seven years of methodological research on how to measure facial behavior. We adopted the particular approach followed in FACS six years ago, and virtually all of our efforts have been expended in the development of this Manual during this time. We know the following about FACS:

People without any scientific training can learn FACS.

FACS can be learned in about 100 hours, taking five weeks if 3-4 hours is spent per day.

People can learn to use FACS without any direct contact with us .

Reliability is satisfactory, with high agreement among the learners and high agreement with how we score facial action. (See Chapter 2 on reliability.)

FACS allows the description of all facial behavior we have observed, and every facial action we have attempted.

FACS successfully postdicts observers' judgments of emotions shown in films or photographs, and a poser's intent about what he is showinq.

Currently, we are using FACS to study conversational facial signals (movements which illustrate speech much as hand movements can) , changes in facial action over the course of hospitalization for depression, and differences between honest and deceptive facial behavior. FACS is proving workable for each of these applications. A number of people whom we have trained in the use of FACS have found it useful in measuring facial behavior in a variety of applications: study of neonatal facial behavior; facial movement shown by deaf persons when they engage in sign language; measuring facial movement changes in response to pharmacological treatment of retarded children; changes in facial behavior associated with changes in heart rate, GSR, and other psychophysiological measures.

We do not have complete information about certain methodological issues:

Is the reliability obtained after learning FACS maintained over a long period of time, or is there a need to refresh or retrain to sustain reliability? We know that without any use of FACS over a period of a few months reliability is maintained when scoring is resumed. We do not know how well reliability is maintained with continued use of FACS over a period of six months to a year.

How many hours per day can someone use FACS before reliability falls off? We suspect it is about three hours a day.

Are all the rules provided in FACS to enhance reliability necessary? Are additional rules needed?

Are the settings or calibrations of the requirements for scoring minimum changes in appearance optimal, too high or too low? We suspect many can be adjusted to pick up more subtle signs of action.

Are all the distinctions in appearance which FACS allows needed for any study of facial activity? We believe the answer will be positive for certain studies.

Answers to these questions requires more experience using FACS to measure various samples of facial behavior. If those who use FACS share with us their experience and problems, their answers to the questions listed above, we will compile their experiences with ours, and determine what changes are needed. We invite you to write us about your experience. We will then revise FACS and share that information with you.