FACIAL EXPRESSION. See Nonverbal Communica- ness). The uniqueness can be further broken down into tion. specific variance (the systematic variance unique to that particular measure) and random error variance. The above equation represents the structure of the measured variables as a function of the underlying la- FACTOR ANALYSIS 1-efers to a broad family of mul- tent factors. Alternatively, the correlational structure tivariate linear model:; designed to examine the inter- among the measured variables can be derived from the relations among a set of continuously distributed man- equation in matrix terms such that ifest variables as a function of a smaller set of + R = A@h' 0 unobserved latent fac~orsT. his technique is often used to examine the common influences believed to give rise where R is the p X p symmetric correlation matrix of to a set of observed measures (measurement structure) p-measured variables, A is the p X rn matrix of factor or to reduce a larger set of measures to a smaller set loadings h, @ is the rn X rn symmetric correlation ma- of linear composites for use in subsequent analysis trix of the latent factors, and 0 is the p x p diagonal (data reduction). Spearman first introduced factor anal- matrix of unique variances E. If @ is a diagonal matrix ysis as a method for mderstanding the common influ- of ones (an identity matrix), then the factors are or- ences thought to underlie multiple measures of intel- thogonal, or uncorrelated. If, instead, @ is a symmetric ligence (I 904). and Thurstone extended Spearman's correlation matrix, then the factors are oblique, or cor- model with the development of the common factor related. The matrix of factor loadings A is called the model (1935). factor pattern matrix, and the matrix of zero-order cor- Factor analysis is based on the premise that the mea- relations between the measures and the factors is called sured variables are a linear additive function of the the factor structure matrix. The factor pattern and fac- unobserved latent factors, and that these factors give tor structure matrices are equal only if the factors are rise to the pattern oi observed correlations among the orthogonal: if the factors are oblique the factor pattern measures. The fundamental equation for the factor matrix is usually consulted for purposes of interpreta- model is: tion. There are two related approaches to the factor Y, = h,151 -!r h,LS* + . . . +h,m5m+ E, model, component analysis and common factor analy- where represents )bej tll of p continuously distributed sis. Component analysis is not a true factor analytic y, variables measured on a sample of n independent sub- model because the resulting components are direct lin- jects. A,,,,r epresents the factor loading (or partial re- ear combinations of the measured variables. However, gression coefficient) relating variable j to the inth factor it is often considered with the common factor model, <,,, and E, represents the influence that is unique to given several shared similarities. In component analy- variable y, and is independent of all !$ and all other ES. sis, it is assumed that all variables are measured with- The variance of the observed measure is thus an ad- out error and that all observed variance among the ditive combination of the variance associated with the measures is available for factoring. Thus, 0 is set to zero set of underlying factors (the communality) and the and the full correlation matrix Rf (containing values of variance associated with the unique factor (the unique- ones on the diagonal) is factored. In contrast, the com- 313 314 FACTOR ANALYSIS mon factor model assumes that the observed variables placement on the diagonal of the reduced correlation are measured with error and that less than the ob- matrix R,. Methods for estimating communalities typi- served variance is available for factoring. Thus, 0 is not cally utilize an initial estimate that may then be itera- constrained to zero and the reduced correlation matrix tively updated during the computation of the factor so- R, (containing values less than one on the diagonal) is lution. Initial estimates include the largest correlation factored. The diagonal elements of R, are communality of each variable with all other variables and the estimates and represent the proportion of variance of squared multiple correlation of each measure with all each measured variable shared with the set of under- other measures. The method used for communality es- lying factors. timation becomes less important given larger numbers of measured variables. Basic Steps in Factor Analysis Factor Rotation. The factors extracted from R, can Once an appropriate sample and set of measures have be rotated to aid in substantive interpretation. Factor been obtained, there are generally four major decision rotation is possible because, for any one factor solution points when using factor analytic models: method of that fits the data to a specific degree, there will exist an factor extraction, number of factors to extract, com- infinite number of equally good solutions, each repre- munality estimation, and factor rotation. sented by a different factor loading matrix. Rotations Factor Extraction. One method of factor extraction can be orthogonal (e.g., varimax, quartimax, and equi- used for both the component and the common factor max). in which the rotated factors are uncorrelated, or model is the method of principal factors. Principal fac- oblique (e.g., promax, direct oblimun, and orthoblique), tors is based on the computation of the eigenvalues and in which the rotated factors are correlated. The goal of vectors (or characteristic roots and vectors) of the cor- factor rotation is to achieve Thurstone’s simple struc- relation matrix to maximize the variance of each suc- ture and the selection of an appropriate rota- (1935), cessively extracted factor. Principal factoring can be ap- tion is usually based on theory and the interpretability plied either to the full correlation matrix Rr that results of the resulting solution. in principal components analysis, or to the reduced cor- Factor Score Estimation. Once the factor analysis relation matrix R, that results in principal axis factor- is completed, factor scores can be estimated that rep- ing. A more recently developed and commonly used resent the scores that would have been observed for an technique for factoring R, is maximum likelihood esti- individual if the latent factors could be measured di- mation, in which model parameters are estimated with rectly. The common factor model is considered indeter- the highest likelihood of having produced the observed minate because there are more model parameters es- correlation matrix. Additional extraction techniques in- timated than pieces of information observed. Thus, clude minimum residual analysis, alpha factoring, and factor scores cannot be computed directly (as is possible image analysis. in component analysis), but must be estimated using Number of Factors. The next step is to determine methods such as the regression method, the Bartlett the number of factors to be extracted. The Kaiser- method, the Anderson-Rubin approach, or simple unit Guttman Rule suggests that the number of factors weighting. Factor scores can then be used in subse- should correspond to the number of eigenvalues of the quent analysis, such as multiple regression or MAN- full correlation matrix Rr that exceed one. although this OVA. criteria is criticized given the tendency to overextract Variants of the Factor Model factors. The scree plot is a graph of each eigenvalue plotted in descending order, and the number of factors The models described above are often termed explora- is determined at the point where an appreciable “bend” tory factor analysis (EFA) and are applied when strong occurs in the plot. The maximum likelihood goodness- theory is lacking and the observed data are freely ex- of-fit test evaluates the magnitude of the residuals that plored in search of meaningful patterns among the ob- exist in R after extracting a given number of factors, servations. In contrast, confirmatory factor analysis and it tests whether additional factors are necessary to (CFA) provides formal statistical tests of a priori hy- meaningfully reduce the size of the residuals. Finally, potheses about the specific factor structure thought to cross-validation can be used in which two independent underlie the set of observed measures (Jiireskog, 1969). samples with two sets of measures from the same do- Unlike the EFA model, in which all measured variables main are examined simultaneously to evaluate the op- relate to all latent factors, the CFA model imposes ex- timal number of factors to extract from each sample. plicit restrictions on the factor pattern matrix so that Determining the appropriate number of factors to ex- the measured variables relate with some (or usually just tract is often thought to be the most important decision one) latent factors but do not relate with others. Test in factor analysis. statistics and goodness-of-fit indices evaluate the ade- Communality Estimation. For the common factor quacy of these imposed restrictions as a function of the model, communality estimates must be obtained for observed sample data. One type of structural equation FALSE MEMORY 315 model is a ('FA model in which one or more of the text that addresses a wide variety of latent variable latent factors are specified as dependent measures in a models, including exploratory factor analysis, confirm- atory factor analysis, and structural equation models. system of regression equations. Many applied examples and discussions of computer There are many techniques related to the factor programs which greatly aid in practical understanding. model that share the general goal of examining simi- Mulaik, S. A. (1972). Thefoundations of factor analysis. New larities among observations. Correspondence analysis is York: McGraw-Hill. A comprehensive advanced math- a nonlinear principal-components analysis of raw cat- ematical treatment of the factor analytic model from a egorical response data, or co-occurrence matrices of rigorous matrix algebra approach. Provides several de- paired preferences. Prolile analysis is used when a set tailed chapters reviewing the basic mathematical foun- of measures is gathered from multiple groups of indi- dations of factor analysis, including matrix algebra, viduals and the similarity of profiles of means is ex- composite variables, and partial correlation. A classic amined across the groups. Cluster analysis classifies text in the general factor model. previously unclassified observations into discrete Spearman, C. (1904). General intelligence. objectively de- termined and measured. American Journal of Psychology. groups based on two or more distance or similarity measures among the observations. Multidimensional 15, 201-293. Thurstone. L. 1,. (1935). The vectors of the mind. Chicago: scaling graphically maps a set of observed distance or University Chicago Press. similarity measures bet ween pairs of items onto one or more underlying dimension, and multidimensional un- Patrick J. Curran folding IS a method for scaling individual preferences among a set of ranked stimuli. FALSE MEMORY. False memories occur when people Bibliography remember events differently from the way they hap- pened or, in the most extreme case, remember events Bollen, K. A. (1989).S tructural equations with latent varia- that never happened at all. These memory errors differ bles. New York: Wiley. The definitive advanced from the most studied types of memory errors, those graduate-level text for matrix-based confirmatory factor involving forgetting. Forgetting is an error of omission, analysis and structural equation modeling. Difficulty arising when a person tries to remember some bit of level is moderate to high. A revision from the same prior knowledge (a name, a fact. a previous occur- publisher is forthcoming. rence), but it does not come to mind. False memories Comrey. A. & Lee, H. B. (1992). A first course in factor L, analysis (2nd ed.). Hillsdale, NJ: Erlbaum. A compre- are errors of commission, because details, facts, or hensive introduction to exploratory factor analysis with events come to mind, often vividly, but the remem- chapters addressing confirmatory factor analysis and brances fail to correspond to prior events. structural equation modeling. Provides a detailed ex- The vicissitudes of memory were studied by Euro- ample of factor analysis as applied to the development pean psychologists near the beginning of the twentieth of the Comrey Personality Scales. century, especially by William Stern in Germany and (;orsuch. R. 1,. (1983). Factor analysis (2nd ed.). Hillsdale, Alfred Binet in France, but the first landmark study was N]: Erlbaum. A widely used graduate-level introduction by Frederic Bartlett, a British psychologist. In his fa- to exploratory factor analysis for the behavioral sci- mous book Remembering Bartlett emphasized ences. The text is moderately mathematical but is ac- (I932), the constructive nature of memory. He argued that rec- cessible to a varie:y of readers. Discusses both concep- ollection of experiences is a reconstructive process tual and practical issues in factor analysis as well as some discussion of computer programs. driven by schemas, or general organizational schemes. Joreskog.K . C. (T969).A general approach to confirmatory The basic idea is that specific experiences and their de- maximum likelihood factor analysis. Psychometrika, 34, tails may not be remembered, but overall themes are. I x j-202. When people try to recover memories, they are guided Kim. J. O., 8r Mueller, C. W. (1978).F actor analysis: Statis- by these schemas or themes and fill in details that are tical rnrthods and practical issues. Beverly Hills, CA: Sage. consistent with the schemas. but that may actually be Kim, J. 0.. & Muelier. C. W. (1978). Introduction to factor quite wrong. malqsis: What it is arid how to do it. Beverly Hills, CA: A key concept to understanding false memories is Sage. Both Kim & Mueller guides provide a brief non- recoding. People do not directly record experience as a technical introduction to the basic concepts and prac- faithful copy of the outside world, but rather recode it tical application of exploratory factor analysis. Also in terms of their knowledge of the world (their sche- contains a short but useful glossary of factor analytic terms and frequently asked questions. mas). In one famous experiment, people looked at am- Loehlin. J. C. (1998).L atent variable models: An introduction biguous shapes and were told to remember them. For to frrc'tor. path, and structural analysis (3rd ed.). Hillsdale, two groups of people, the shapes were given different K J: Erlbaum. An excellent graduate-level introductory labels (such as, for one shape, either broom or rifle). FALSE MEMORY 315 model is a ('FA model in which one or more of the text that addresses a wide variety of latent variable latent factors are specified as dependent measures in a models, including exploratory factor analysis, confirm- atory factor analysis, and structural equation models. system of regression equations. Many applied examples and discussions of computer There are many techniques related to the factor programs which greatly aid in practical understanding. model that share the general goal of examining simi- Mulaik, S. A. (1972). Thefoundations of factor analysis. New larities among observations. Correspondence analysis is York: McGraw-Hill. A comprehensive advanced math- a nonlinear principal-components analysis of raw cat- ematical treatment of the factor analytic model from a egorical response data, or co-occurrence matrices of rigorous matrix algebra approach. Provides several de- paired preferences. Prolile analysis is used when a set tailed chapters reviewing the basic mathematical foun- of measures is gathered from multiple groups of indi- dations of factor analysis, including matrix algebra, viduals and the similarity of profiles of means is ex- composite variables, and partial correlation. A classic amined across the groups. Cluster analysis classifies text in the general factor model. previously unclassified observations into discrete Spearman, C. (1904). General intelligence. objectively de- termined and measured. American Journal of Psychology. groups based on two or more distance or similarity measures among the observations. Multidimensional 15, 201-293. Thurstone. L. 1,. (1935). The vectors of the mind. Chicago: scaling graphically maps a set of observed distance or University Chicago Press. similarity measures bet ween pairs of items onto one or more underlying dimension, and multidimensional un- Patrick J. Curran folding IS a method for scaling individual preferences among a set of ranked stimuli. FALSE MEMORY. False memories occur when people Bibliography remember events differently from the way they hap- pened or, in the most extreme case, remember events Bollen, K. A. (1989).S tructural equations with latent varia- that never happened at all. These memory errors differ bles. New York: Wiley. The definitive advanced from the most studied types of memory errors, those graduate-level text for matrix-based confirmatory factor involving forgetting. Forgetting is an error of omission, analysis and structural equation modeling. Difficulty arising when a person tries to remember some bit of level is moderate to high. A revision from the same prior knowledge (a name, a fact. a previous occur- publisher is forthcoming. rence), but it does not come to mind. False memories Comrey. A. & Lee, H. B. (1992). A first course in factor L, analysis (2nd ed.). Hillsdale, NJ: Erlbaum. A compre- are errors of commission, because details, facts, or hensive introduction to exploratory factor analysis with events come to mind, often vividly, but the remem- chapters addressing confirmatory factor analysis and brances fail to correspond to prior events. structural equation modeling. Provides a detailed ex- The vicissitudes of memory were studied by Euro- ample of factor analysis as applied to the development pean psychologists near the beginning of the twentieth of the Comrey Personality Scales. century, especially by William Stern in Germany and (;orsuch. R. 1,. (1983). Factor analysis (2nd ed.). Hillsdale, Alfred Binet in France, but the first landmark study was N]: Erlbaum. A widely used graduate-level introduction by Frederic Bartlett, a British psychologist. In his fa- to exploratory factor analysis for the behavioral sci- mous book Remembering Bartlett emphasized ences. The text is moderately mathematical but is ac- (I932), the constructive nature of memory. He argued that rec- cessible to a varie:y of readers. Discusses both concep- ollection of experiences is a reconstructive process tual and practical issues in factor analysis as well as some discussion of computer programs. driven by schemas, or general organizational schemes. Joreskog.K . C. (T969).A general approach to confirmatory The basic idea is that specific experiences and their de- maximum likelihood factor analysis. Psychometrika, 34, tails may not be remembered, but overall themes are. I x j-202. When people try to recover memories, they are guided Kim. J. O., 8r Mueller, C. W. (1978).F actor analysis: Statis- by these schemas or themes and fill in details that are tical rnrthods and practical issues. Beverly Hills, CA: Sage. consistent with the schemas. but that may actually be Kim, J. 0.. & Muelier. C. W. (1978). Introduction to factor quite wrong. malqsis: What it is arid how to do it. Beverly Hills, CA: A key concept to understanding false memories is Sage. Both Kim & Mueller guides provide a brief non- recoding. People do not directly record experience as a technical introduction to the basic concepts and prac- faithful copy of the outside world, but rather recode it tical application of exploratory factor analysis. Also in terms of their knowledge of the world (their sche- contains a short but useful glossary of factor analytic terms and frequently asked questions. mas). In one famous experiment, people looked at am- Loehlin. J. C. (1998).L atent variable models: An introduction biguous shapes and were told to remember them. For to frrc'tor. path, and structural analysis (3rd ed.). Hillsdale, two groups of people, the shapes were given different K J: Erlbaum. An excellent graduate-level introductory labels (such as, for one shape, either broom or rifle). 316 FALSE MEMORY When these people were later asked to draw the shapes “the truth” of the original situation, experimental evi- from memory, they tended to draw them to be consis- dence shows that they can be wrong. In fact, some tent with the label. The ambiguous figure called a studies show little correlation between confidence and broom during the study was recalled as being more accuracy in memories. Many people who are cautious broomlike than it was originally. In this case, people about their prowess in remembering may actually be recoded their visual experience in terms of language more accurate than others who are perpetually sure and then, apparently, remembered the recoded form. that they are right. Such processes occur in virtually all types of remem- Guessing and imagining can lead to false memories, bering: People do not recollect the events of the world too. If people are first tested in a situation in which as they actually happened, but as they are recoded by they are told to guess about what might have happened the cognitive system. (an instruction often given under conditions of hyp- The coherence or interrelatedness of presented in- nosis), they often produce wrong information. Later, formation is another factor that can lead to false mem- when trying to remember accurately, they may recall ories. When what one reads or hears strongly implies events that they had produced while guessing. In a a fact or event, people can come to believe that the sense, these people have provided their own misinfor- event occurred, even when it did not. For example, mation about the earlier events and now remember when a person hears, “The karate champion hit the their self-created rendition of the events. Imagination cinder block,” that person may later remember having can also lead to false memories. If people repeatedly heard that the karate champion broke the cinder block. imagine acting a certain way in a particular situation. Breaking the block may be implied, but is not stated they can come to believe that the events actually hap- and may not have occurred (the champion might have pened, especially when the imaginings fit plausibly with failed). As people perceive the world around them, they a person’s prior experience. make inferences about what the events mean, and may These cases all fall under the rubric of what has later confuse their inferences (the champion broke the been called reality monitoring or source monitoring: Peo- block) with what was actually stated (the champion hit ple confuse what actually happened (reality) with what the block). Again, we remember our recoded experi- they may have only inferred, imagined, or heard about. ences, not the literal happenings of the outside world. When people remember details provided later as having A number of factors aside from recoding lead to false actually been present in the original event, they are memories. Interference processes are responsible for combining information from two sources. In these one large class of cases. One popularly studied situation cases, false memories arise due to a failure to monitor involves eyewitness testimony, modeled after cases that the sources of information. might arise in legal settings. A person witnesses a traf- Individual differences among people are also an im- fic accident with a yield sign present. Later, the person portant consideration. Some studies find that children reads a description of the accident in which most of seem especially prone to false recollections relative to the information is accurate, but a stop sign is men- young adults. Older adults may also be more susceptible tioned rather than a yield sign (i.e., misleading infor- than are young adults. One neuropsychological ac- mation is provided). When people were later asked to count of these differences emphasizes brain develop- recall or to recognize the original accident scene, they ment, particularly frontal lobe functioning. Certain often remembered the sign as having been a stop sign areas of the frontal lobes are believed to be critical in rather than a yield sign, relative to recall or recognition memory retrieval and in monitoring the source of in- in control conditions in which no misleading informa- formation. Children’s frontal lobes are still developing, tion had been given. Reading about the stop sign cre- and in older adults these areas atrophy, perhaps leading ated interference, which caused people to remember the to the exaggerated proclivity to false recollections in wrong sign when they tried to recollect the original these two groups. In addition, some individuals with scene. The critical point is that our memories for some severe frontal damage display rather dramatic false events are not encapsulated in time and impervious to memories, including confabulation (recollections of further manipulation, but can be changed by the pro- wildly implausible events asserted as fact). The tie be- cessing of subsequent information. tween frontal functioning and the arousal of false Social factors can also affect remembering and cause memories is still being actively investigated. false recollections. When people remember in groups, The issue of the truth or falsity of memories arises if one person recollects information that is wrong, oth- in many settings outside the psychology lab, particu- ers can incorporate the wrong information into their larly in legal cases. Three types of cases that hinge on memories for the scene. As in the eyewitness paradigm, accuracy of remembering will be discussed in turn. The the misleading information provided by one person can first is the general issue of the accuracy of eyewitness distort the memories of others in the group. Even when memory of a crime. say with regard to identifying a people are highly confident that they are remembering suspect as the criminal. Such testimony carries great FAMILY PSYCHOLOGY: History of the Field 317 weight with juries-having a witness say, "That is the completely erroneous, hence simple recollection should man. 1'11 never forget his face!"-seems very compel- not be routinely accepted as completely accurate. ling. However. as discussed before, even highly confi- [See also Repressed Memory.] dent witnesses can sometmes be completely wrong. so such eyewitness testimony must be regarded cautiously, unless corroborated by other evidence. In addition, be- Bibliography cause memories can be .,nfluenced by suggestion, the means of elicitation of the eyewitness memories must Bartlett. F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge, England: Cambridge be examined carefully. University Press. A second type of case involves the child witness. Brewer, W. F. (1977). Memory for the pragmatic implica- Some adults (teachers, clay-care operators) have been tions of sentences. Memory and Cognition, 5, 673-678. accused of committing crimes of childhood sexual Carmichael, L. L., Hogan, H. P.. & Walter, A. A. An (1932). abuse based only on ch.ildren's testimony. Of course, experimental study of the effect of language on repro- such crimes are horrifi: and perpetrators should be duction of visually perceived form. Journal of Experi- brought to justice. On the other hand, false accusations mental Psychology, 15, 73-8 j. can ruin innocent people's lives. Because children seem Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A quite susceptible to crea.ted memories through sugges- scientific analysis of children's testimony. Washington, tive questioning techniques, it is again critical to ask DC: American Psychological Association. how the memories wen3 elicited and whether there is Jacoby, L. L., Kelley, C.. Brown, J., & Jasechko. J. (1989). Becoming famous overnight: Limits on the ability to strong corroborating evidence from other sources. avoid unconscious influences of the past. Journal of Per- A third type of case involves the issue of recovered sonality and Social Psychology, 56, 326-338. memories after long delays in adults. In recent times, Johnson, M. K., & Raye. C. L. (1981). Reality monitoring. many cases have occurred in which adults seem to re- Psychological Review, 88, 67-85. cover memories of haking been physically or sexually Loftus, E. F. (1993). The reality of repressed memories. abused as children, with such recoveries often occur- American Psychologist, 48, 518-537. ring while they are undergoing therapy. They often ac- Loftus. E. E. Miller, D. G., & Burns, H. J. (1978). Semantic cuse a parent, another relative, or a teacher of the integration of verbal information into a visual memory. abuse. The general idea behind such recoveries is that Journal of Experimental Psychology: Human Learning and the original events were traumatic and hence repressed, Memory. 4* 19-31. or banished to an unconscious state, from which they Pendergrast, M. (199 5). Victims of memory: Incest accusa- tions and shattered lives. Hinesburg. VT: Upper Access. are later recovered in therapy. A debate rages about the Roediger, H. L.. 111. (1996). Memory illusions. Journal of veracity of these recovered memories. Are they real? Or Memory and Language, 35, 76-100. are they false memories aroused through suggestive Roediger, H. L., 111, & McDermott, K. B. (2000). Distortions questioning and through techniques such as hypnosis of memory. In E. Tulving & F. I. M. Craik (Eds.). Oxford and imagination. which are known to lead to false handbook of memory. New York: Oxford University Press. memories? Some of these memories seem so improbable Schacter. D. L., Coyle, J. T., Fischbach, G. D., Mesulam, (ones involving cults practicing satanic ritual abuse, for M. M., & Sullivan, L. E. (Eds.). (1995). Memory distor- which there is no evidence) or violate known facts tion: How minds, brains, and societies reconstruct the past. about memory (reco17ering detailed accounts of abuse Cambridge, MA: Harvard University Press. that supposedly happened before one year of age, a Henry L. Roediger 111 and David A. Gallo time about which people are known to consciously re- member little or nothing), that there seems little doubt that they are false. Bcdt of course many other recovered memories are at least possible. Once again, the twin issues to raise in these cases are the techniques used to FAMILY PSYCHOLOGY. [This entry comprises three ar- elicit the memories (Were they suggestive?) and the ticles: an overview of the broad history of thefieldfrom its supporting evidence (Is there any!). In general, because inception to the present; a survey of the principal theories the events happened so long ago. it is difficult to find that have determined the course of development of thefield; evidence to corroborate or to disconfirm the memories. and a general descriptive review of assessments and inter- The study of the vagaries of remembering has pro- ventions used in or unique to the field. For a discussion duced an interesting body of scientific information that related to family psychology, see Family Therapy.] is important for forensic purposes. Simply the claim that "I remember" 'should not, without converging ev- History of the Field idence, be taken as an indicator that remembered events actually occurred. Much evidence exists that Family psychology is a specialty within contemporary even detailed. highly confident recollections can be psychology. It has basic and applied aspects, covers FAMILY PSYCHOLOGY: History of the Field 317 weight with juries-having a witness say, "That is the completely erroneous, hence simple recollection should man. 1'11 never forget his face!"-seems very compel- not be routinely accepted as completely accurate. ling. However. as discussed before, even highly confi- [See also Repressed Memory.] dent witnesses can sometmes be completely wrong. so such eyewitness testimony must be regarded cautiously, unless corroborated by other evidence. In addition, be- Bibliography cause memories can be .,nfluenced by suggestion, the means of elicitation of the eyewitness memories must Bartlett. F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge, England: Cambridge be examined carefully. University Press. A second type of case involves the child witness. Brewer, W. F. (1977). Memory for the pragmatic implica- Some adults (teachers, clay-care operators) have been tions of sentences. Memory and Cognition, 5, 673-678. accused of committing crimes of childhood sexual Carmichael, L. L., Hogan, H. P.. & Walter, A. A. An (1932). abuse based only on ch.ildren's testimony. Of course, experimental study of the effect of language on repro- such crimes are horrifi: and perpetrators should be duction of visually perceived form. Journal of Experi- brought to justice. On the other hand, false accusations mental Psychology, 15, 73-8 j. can ruin innocent people's lives. Because children seem Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A quite susceptible to crea.ted memories through sugges- scientific analysis of children's testimony. Washington, tive questioning techniques, it is again critical to ask DC: American Psychological Association. how the memories wen3 elicited and whether there is Jacoby, L. L., Kelley, C.. Brown, J., & Jasechko. J. (1989). Becoming famous overnight: Limits on the ability to strong corroborating evidence from other sources. avoid unconscious influences of the past. Journal of Per- A third type of case involves the issue of recovered sonality and Social Psychology, 56, 326-338. memories after long delays in adults. In recent times, Johnson, M. K., & Raye. C. L. (1981). Reality monitoring. many cases have occurred in which adults seem to re- Psychological Review, 88, 67-85. cover memories of haking been physically or sexually Loftus, E. F. (1993). The reality of repressed memories. abused as children, with such recoveries often occur- American Psychologist, 48, 518-537. ring while they are undergoing therapy. They often ac- Loftus. E. E. Miller, D. G., & Burns, H. J. (1978). Semantic cuse a parent, another relative, or a teacher of the integration of verbal information into a visual memory. abuse. The general idea behind such recoveries is that Journal of Experimental Psychology: Human Learning and the original events were traumatic and hence repressed, Memory. 4* 19-31. or banished to an unconscious state, from which they Pendergrast, M. (199 5). Victims of memory: Incest accusa- tions and shattered lives. Hinesburg. VT: Upper Access. are later recovered in therapy. A debate rages about the Roediger, H. L.. 111. (1996). Memory illusions. Journal of veracity of these recovered memories. Are they real? Or Memory and Language, 35, 76-100. are they false memories aroused through suggestive Roediger, H. L., 111, & McDermott, K. B. (2000). Distortions questioning and through techniques such as hypnosis of memory. In E. Tulving & F. I. M. Craik (Eds.). Oxford and imagination. which are known to lead to false handbook of memory. New York: Oxford University Press. memories? Some of these memories seem so improbable Schacter. D. L., Coyle, J. T., Fischbach, G. D., Mesulam, (ones involving cults practicing satanic ritual abuse, for M. M., & Sullivan, L. E. (Eds.). (1995). Memory distor- which there is no evidence) or violate known facts tion: How minds, brains, and societies reconstruct the past. about memory (reco17ering detailed accounts of abuse Cambridge, MA: Harvard University Press. that supposedly happened before one year of age, a Henry L. Roediger 111 and David A. Gallo time about which people are known to consciously re- member little or nothing), that there seems little doubt that they are false. Bcdt of course many other recovered memories are at least possible. Once again, the twin issues to raise in these cases are the techniques used to FAMILY PSYCHOLOGY. [This entry comprises three ar- elicit the memories (Were they suggestive?) and the ticles: an overview of the broad history of thefieldfrom its supporting evidence (Is there any!). In general, because inception to the present; a survey of the principal theories the events happened so long ago. it is difficult to find that have determined the course of development of thefield; evidence to corroborate or to disconfirm the memories. and a general descriptive review of assessments and inter- The study of the vagaries of remembering has pro- ventions used in or unique to the field. For a discussion duced an interesting body of scientific information that related to family psychology, see Family Therapy.] is important for forensic purposes. Simply the claim that "I remember" 'should not, without converging ev- History of the Field idence, be taken as an indicator that remembered events actually occurred. Much evidence exists that Family psychology is a specialty within contemporary even detailed. highly confident recollections can be psychology. It has basic and applied aspects, covers 318 FAMILY PSYCHOLOGY: History of the Field treatment and prevention interventions, and is taught ical purviews, but it has moved beyond family as well in graduate training programs, generally within clini- (L’Abate, 1992; Parke, 1998). The emphasis is more on cal/child or counseling psychology. Although it has less the notion of behavioral transactions that occur not of a formal and lengthy history than more well known only in the family, but between the family and other and established specialties like clinical, counseling, or developmentally influential social systems (Bronfen- school psychology, it stands among the new specialties brenner, 1979). This broadened emphasis, in accord such as health and forensic psychology. Its name may with Kazdin’s (1997) recommendations for new kinds indicate otherwise, but the specialty of family psychol- of conceptualization and intervention, has been the ogy covers a broad territory (L’Abate, 1985). Advances hallmark of contemporary family psychology interven- covering family and marital issues and processes re- tions (Liddle, 1995; Markman, Tolan, Guerra, & 1992: search and interventions, as well as the core theoretical Kendall, 1995). Peer group and other institutions of in- ideas of family psychology (contextualism and systems fluence beyond the family are included in family psy- perspective) appear under the headings of many psy- chology’s conceptual framework. and today’s interven- chology content areas, including developmental, social, tions that could be said to reside in the family research design and methodology, environmental and psychology tradition include assessment and interven- community, gender, industrial/organizational, health, tion attention to multiple systems and levels of social forensic, and, of course clinical, counseling, and school influence, including peer, school, and community/ psychology (Crosbie-Burnett & Lewis, 1993; Liddle, neighborhood influences (Alexander, Holtzworth- 19874. Munroe, & Jameson, 1994; Henggeler et al., 1997; Lid- One of the first tasks in family psychology’s early dle, 1998; Szapocznik & Coatsworth, in press). history concerned its definition, in particular, its defi- Although the family, marriage, and systems theory nition vis-a-vis the more widely recognized movement, and contextualism-theoretical notions that are core to field, and clinical specialty known as family therapy family psychology-have been of interest to psycholo- (Kaslow, 1991; L’Abate, 1992). The first issue of the gists for some time, it was with the founding of the Journal of Family Psychology included the following def- Division of Family Psychology in 1985 and 2 years inition: later, its journal, the Journal of Family Psychology (JFP), that the family psychology specialty achieved a well- Family psychology, using a systemic perspective, broad- defined and differentiated identity within mainstream ens psychology’s traditional emphasis on the individual, psychology. The division became a home for interests and, while it retains a primary emphasis on marriage relevant to psychologists who wished to practice from and the family, it uses the systemic view to focus on a family therapy or systems perspective, for those who the nature and role of individuals in primary relation- trained students in family systems ideas, research, and ship structures, as well as, more broadly, the social ecol- interventions, and for those concerned with carrying ogy of the family-those networks in which the family interacts and resides. (Liddle, 1987a. p. 9) the standard of family and couple-related policy issues within the American Psychological Association (APA) Definitions of the specialty emphasized several themes: as well as on the federal policy scene. The founding of the systems perspective and the contextual tradition of JFP (first a divisional journal and now an APA journal) family therapy (representing the clinical wing of family is an endorsement of the specialty’s centrality in con- psychology); the theoretical perspectives of contextual- temporary psychology, and can be considered another ism: reciprocal causality and influence: and an ongoing major event for the specialty. JFP became a context attempt to define the processes or mechanisms that me- through which family psychology, as an identifiable and diate the relationship between different systems and differentiated specialty within psychology, could define levels of functioning (Kaslow, 1995; Liddle, 1987b). itself (Levant, 1992). The range of articles published in Family psychology embraced the newly forming tradi- JFP has been broad, and has included basic normative tion of conceptual complexity, which were taking place and dysfunctional marital and family process research, in many sectors of psychology and related sciences research on diverse family structures and forms, re- (Bell, 1968; Haynes, 1992: Kazdin & Kagan, 1994). One search that informs social and public policy pertaining of the most important ways in which family psychology to families, and treatment outcome and process studies. exemplified and, in its own way, developed along these Several other important developments have occurred in complex lines, concerned the way it went beyond the the dozen or so years since the founding of the Division literal meaning of the very term family psychology. Fam- of Family Psychology and the specialty’s journal. ily psychology, like family therapy, invokes images of a Training Programs and Accreditation particular conceptual or intervention unit-the family. Family psychology, embracing systems philosophy and Ideas about training and supervision that are syntonic contextualism at its core, surely includes family and with systemic principles have been articulated (Ganahl, marital processes in its theoretical, research, and clin- Ferguson, & L‘Abate, 1985; Green, 1998; Liddle. Becker, FAMILY PSYCHOLOGY: History of the Field 319 & Ijiamond, ~997)C. omplex, clearly specified doctoral are being conducted to test a new generation of inter- level training models and programs have been ad- ventions that are comprehensive in scope, target indi- vanced as well (Berger, 1988; Green, 1998). Programs vidual and family processes and family vis-a-vis extra- such as these await the accreditation of family psy- familial interactional processes (Schoenwald & chology as a formal specialty in the APA, a process that Henggeler, in press). In conclusion, advances on many has begun and seems likely. fronts-scientific, clinical, theoretical, and organiza- Family psychology ideas are now developed within tional-are evident in family psychology. It continues conceptual models in and of themselves, but are also to be defined as what has been called an emerging and part of other evolving models of theory and practice. an emerged discipline (Liddle, 1987b). Many aspects of the family psychology research, theory, and clinical practice agenda of over a decade ago (Lid- dle, 1987b) have non been reached. Family psychology Bibliography has proven to be a viable specialty within contemporary psychology. It has been an organizing influence for re- Alexander, J. F., Holtzworth-Munroe, A., & Jameson, P. B. search on systemic theory and ideals within psychol- (1994). The process and outcome of marital and family therapy: Research review and evaluation. In A. E. Ber- ogy: it has developed and rigorously tested new, family gin & S. I. Garfield (Eds.),H andbook of psychotherapy and psychology theory-based (e.g., using developmental re- behavior change (4th ed., pp. 595-630). New York: Wiley. search as a guide to practice) preventive and treatment Bell, R. Q. (1968). A reinterpretation of the direction of interventions: and it has been an active participant in effects in studies of socialization. Psychological Review, legislation and health/mental health policy making. 75 (21, 81-95. Berger, M. (1988). Academic psychology and family ther- Research apy training. In H. A. Liddle & D. C. Breunlin (Eds.), The research landscape changed dramatically in family Handbook of family therapy training and supervision psychology in the 1g3os. Many of those changes be- (pp. 303-315). New York: Guilford Press. came apparent at the first national conference on fam- Bray, J. (in press). Methodological issues and innovations in family psychology intervention research. In H. A. ily psychology, sponsored by the APA Science Director- ate in 1995 (Liddle, Santisteban, Levant, & Bray, in Liddle, D. A. Santisteban, R. Levant, & J. Bray (Eds.), Family psychology intervention science. Washington, DC: press). Long-term stahle programs of basic and inter- American Psychological Association. vention (prevention and treatment) research exist (Sza- Bronfenbrenner, U. (1979). The ecology of human develop- pocznik. Kurtines, Santisteban, & Rio, 1990). A strong ment: Experiments by nature and design. Cambridge, MA: interest in family and marital processes in both the ap- Harvard University Press. plied and basic science areas can be found across fed- Crosbie-Burnett, M., & Lewis, E. (1993). Theoretical con- eral. state. and foundation funding sources. More fund- tributions from social and cognitive-behavioral psy- ing, from a wide variely of agencies (National Institute chology. In €? G. Boss, W. J. Doherty, R., LaRossa, W. R. of Mental Health. National Institute on Drug Abuse, Schumm, & K. Steinmetz (Eds.). Sourcebook of family S. National Institute on Alcohol Abuse and Alcoholism, theories and methods: A contextual approach (pp. 53~- Center for Substance Abuse Treatment, Center for Sub- 561). New York: Plenum Press. Ganahl. G., Ferguson, L. R., & L'Abate, L. (1985). Training stance Abuse Prevention, Centers for Disease Control, in family therapy. In L. L'Abate (Ed.), The handbook of Department of Justice /Office of Juvenile Justice and family psychology and therapy (pp. 128I-T317). Home- Delinquency Prevention) specifically encouraging the wood, IL Dorsey Press. evaluation of family-related intervention initiatives, is Green, R. J. (T998). Training programs: Guidelines for mul- available now than ever before. The National Institute ticultural transformation. In M. McGoldrick (Ed.), Re- on Drug Abuse, for example, in an era in which the visioning family therapy: Race, culture, and gender in clin- overall number of research centers is being reduced, ical practice (pp. 111-117). New York: Guilford Press. recently funded a research center in the area of family- Haynes, (1992). Models of causality in psychopathology: S. based treatments for adolescent substance abuse (Lid- Toward dynamic synthetic and nonlinear models of behav- dle. ~998)F.a mily and couple interventions have been ior disorders. Needham Heights, MA: Allyn & Bacon. developed and tested with a wide variety of clinical Henggeler, S. W., Rowland, M. D., Pickrel, S. G., Miller, problems and populations (Pinsof & Wynne, 1995). The S. I*., Cunningham, €? B., Santos, A. B., Schoenwald, S. K., Randall. J., & Edwards, J. E. (1997). Investigating technology of family psychology intervention science, family-based alternatives to institution-based mental as a subspecialty. has evolved a great deal as well (Bray, health services for youth: Lessons learned from the pilot in press: Snyder, Cozzi, Mangrum, in press). Complex 4% study of a randomized field trial. Journal of Child Clinical interventions have been manualized and tested using Psychology, 26, 226-233. state-of-the-science designs, measures, and statistical Kaslow, F. W. (1991). The art and science of family psy- analyses. New studies, combining different research chology: Retrospective and perspective. American Psy- traditions (effectiveness, efficacy, and process studies) chologist. 46, 621-626. 320 FAMILY PSYCHOLOGY: Theories of Family Dynamics Kaslow, F. W. (Ed.). (1995). Voices in family psychology. Origins and interventions. Washington, DC: American Thousaud Oaks, CA: Sage. Psychological Association. Kazdin, A. E. (1997). A model for developing effective treat- Szapocznik, J., Kurtines W., Santisteban, D. A., & Rio, A. T. ments: Progression and interplay of theory, research, (1990). Interplay of advances between theory, research, and practice. Journal of Clinical Child Psychology, 26, and application in treatment interventions aimed at be- 2 I 7-226. havior problem children and adolescents. Journal of Kazdin, A. E., & Kagan, J. (1994). Models of dysfunction Consulting and Clinical Psychology, 58, 696-703. in developmental psychopathology. Clinical Psychology: Tolan, P. H., Guerra. N. G., & Kendall, €? C. (1995). A Science and Practice, I, 35-52. developmental-ecological perspective on antisocial be- L’Abate, L. (1985). The handbook of family psychology and havior in children and adolescents: Toward a unified therapy (Vols. I & 2). Homewood, IL: Dorsey Press. risk and intervention framework. Journal of Consulting L’Abate. L. (1992). Family psychology and family therapy: and Clinical Psychology, 63, 579-584. Comparisons and contrasts. American Journal of Family Howard A. Liddle Therapy, 20, 3-12. Levant, R. F. (1992). The evolving field and its issues. Jour- nal of Family Psychology, 6, 5-9. Liddle, H. A. (1987a). Editor’s introduction I. Family psy- Theories of Family Dynamics chology: The journal, the field. Journal of Family Psy- Family therapy has multiple origins and it is difficult to chology, I, 5-22. say exactly who conducted the first conjoint family Liddle, H. A. (1987b). Editor’s introduction 11. Family psy- treatment session. One early form of family interven- chology: Tasks of an emerging (and emerged) disci- pline. Journal of Family Psychology, I, 149-167. tion was the settlement house movement in the 1930s Liddle H. A. (1995). Conceptual and clinical dimensions of in which social workers attempted to keep impoverished a multidimensional, multisystems engagement strategy families intact. Child guidance clinics, which originated in family-based adolescent treatment. Psychotherapy: in the early 1900s, included some attention to the role Theory, Research and Practice, 32, 39-58. of parenting. However, in classical child therapy, par- Liddle, H. A. (1998). MultidimensionaI family therapy treat- ents were seen by the social worker, and a psychody- ment manual for the Cannabis Youth Treatment Multisite namically trained psychiatrist worked individually with Collaborative Project. Rockville, MD: Center for Sub- the child. Parents were often seen as causing the child’s stance Abuse Treatment. difficulties, but could be harmful contaminants to in- Liddle, H. A., Becker, D., & Diamond, G. M. (1997). Family dividual therapy. Another early influence was the psy- therapy supervision. In C. E. Watkins (Ed.), Handbook of chiatrist Nathan Ackerman, who noted that fathers’ psychotherapy supervision (pp. 400-418). New York: Wi- ley. unemployment triggered family-wide distress emerging Liddle, H. A., Santisteban. D. A., Levant, R., & Bray, J, in the children’s mental health problems. (Eds.). (in press). Family psychology intervention science. In the early 1950s. a small group of investigators Washington, DC: American Psychological Association. including Gregory Bateson (an anthropologist), Jay Ha- Markman, H. J. (1992). Marital and family psychology: ley (with an academic background in library science Burning issues. Journal of Family Psychology, 5, 264- and communications), John Weakland (an engineer 275. trained in anthropology) and Don Jackson (a psychia- Parke, R. D. (1998). Editorial. Journal of Family Psychology. trist and the only trained mental health professional in 12, 3-6. the group), were studying human and animal com- Pinsof, W., & Wynne, L. (1995).T he efficacy of marital and munication in several contexts. These early family ther- family therapy: An empirical overview, conclusions, apists were often outsiders and either did not agree and recommendations. Journal of Marital and Family Therapy, 21, 585-613. with or were not trained in conventional individual psy- Schoenwald, S., & Henggeler, S. (in press). Services re- chopathology or treatment. Bateson and colleagues ob- search and family-based treatment. In H. A. Liddle, served that young adults with schizophrenia often ex- D. A. Santisteban, R. Levant, & J. Bray (Eds.), Family hibited acute symptom exacerbation immediately psychology intervention science. Washington, DC: Amer- following a family visit. The interactions themselves in- ican Psychological Association. cluded conflicting communication sequences termed Snyder, D. K.. Cozzi, J. C.. & Mangrum. L. F. (in press). Con- double binds. For example, a mother would come to visit ceptual issues in assessing couples and families. In her adult son in the psychiatric unit. The son would H. A. Liddle, D. A. Santisteban, R. Levant, & J, Bray stand at a respectful distance and say, “Hello, mother.” (Eds.), Family psychology intervention science. Washing- She would respond by saying, “Don’t you love your ton, DC: American Psychological Association. mother? Come and give me a hug.” When he embraced Szapocznik, J., & Coatsworth, J.D. (in press). An ecode- velopmental framework for organizing the influences her, she would physically stiffen, pull away, and say: on drug abuse: A developmental model of risk and pro- “Son, you must act more grown up. You’re not a baby tection. In M. Glantz & C. R. Hartel (Eds.), Drug abuse: anymore” (Nichols & Schwartz, 1998). These contra-