Autocorrelation Effects on Least-Squares Intervention Analysis of Short Time Series

Several issues regarding the effects of autocorrelated errors on Type I error in ordinary least-squares models are clarified. Although autocorrelated errors have a large effect on both omnibus F tests and tests on individual intervention effect coefficients in many applications, there are exceptions that have not been pointed out previously. It is demonstrated that under certain conditions, distortion in Type I error is far less than is predicted by asymptotic theory. It is shown that these exceptions occur because the effect of autocorrelated errors is dependent on (a) the type of parameters (e.g., level change and/or slope change) required in the model, (b) the number of variables in the design matrix, and (c) the sample size. Because existing time-series methods perform poorly with small samples, this may be a useful finding in some situations; however, a better general solution is to use a recently developed small-sample method.

[1]  Douglas A. Hibbs,et al.  Problems of Statistical Estimation and Causal Inference in Time-Series Regression Models , 1973 .

[2]  William D. Berry,et al.  New Tools for Social Scientists: Advances and Applications in Research Methods , 1987 .

[3]  J. McKean,et al.  The Runs Test for Autocorrelated Errors: Unacceptable Properties , 1996 .

[4]  W. Agras,et al.  Behavior modification of anorexia nervosa. , 1974, Archives of general psychiatry.

[5]  John M. Gottman,et al.  Time Series Analysis: A Comprehensive Introduction for Social Scientists. , 1983 .

[6]  T. Cook,et al.  Quasi-experimentation: Design & analysis issues for field settings , 1979 .

[7]  Joseph W. McKean,et al.  Irrelevant autocorrelation in least-squares intervention models. , 1998 .

[8]  S. Nash,et al.  Numerical methods and software , 1990 .

[9]  Leonard H. Epstein,et al.  Music feedback in the treatment of tension headache: An experimental case study , 1974 .

[10]  Terry M. Mayhew,et al.  The Design and Analysis of Research Studies , 1993 .

[11]  Thomas R. Kratochwill,et al.  Single Subject Research: Strategies for Evaluating Change , 1978 .

[12]  Charles W. Ostrom.  Time Series Analysis , 1990 .

[13]  G. Glass,et al.  Design and analysis of time-series experiments , 1975 .

[14]  R. Blunden,et al.  Improving Environments for Profoundly Handicapped Adults , 1980 .

[15]  J. McKean,et al.  A double bootstrap method to analyze linear models with autoregressive error terms. , 2000, Psychological methods.

[16]  Charles W. Ostrom.  Time Series Analysis: Regression Techniques , 1978 .

[17]  Joseph W. McKean,et al.  Autocorrelation estimation and inference with small samples. , 1991 .

[18]  John M. Gottman,et al.  4 – Analysis of Interrupted Time-Series Experiments , 1978 .

[19]  J. Durbin,et al.  Testing for serial correlation in least squares regression. I. , 1950, Biometrika.

[20]  R. Liberman,et al.  RESEARCH DESIGN FOR ANALYZING DRUG‐ENVIRONMENT‐BEHAVIOR INTERACTIONS , 1973, The Journal of nervous and mental disease.

[21]  M. Lewis-Beck.  Interrupted Time Series , 1986 .

[22]  J. Durbin,et al.  Testing for serial correlation in least squares regression. II. , 1950, Biometrika.

[23]  D. Barlow,et al.  Single Case Experimental Designs: Strategies for Studying Behavior Change , 1976 .

[24]  R. Jaeger,et al.  Complementary Methods for Research in Education , 1988 .

[25]  A. Mood.  The Distribution Theory of Runs , 1940 .