Measurement biases in user involvement research

Considerable research has been conducted to demonstrate user involvement's effect on information system success. User involvement and system success typically have been measured by asking users for their perceptions of these variables. This paper reports on a field study conducted to investigate the possibility that this approach to measuring study variables tends to overstate the benefits of user involvement. The link of user involvement to user satisfaction is found to be significantly weaker when user involvement is assessed by systems analysts than when it is self reported. Further, this difference is found to be greater for systems with few users than for systems with many users. The findings suggest that common method variance and self-serving bias may have overstated the apparent benefits of user involvement in past research on information systems. Suggestions for future research are presented.     

RENEWAL COUNTING PROCESS INDUCED BY A DISCRETE MARKOV CHAIN

Some renewal theoretic properties of a renewal counting process induced by a Markov chain on the set of non-negative integers are established, namely, analogues of the classical elementary, Blackwell, and Breiman theorems and the key renewal theorem. These results generalize those of Vere-Jones (1975) who considered a Markov chain on the set of positive integers.     

A note on shrinkage factors in two stage testimation

We obtain a simple and natural testimator which has locally, at the parametric point corresponding to the prior knowledge, a smaller mean squared error than any other two stage testimator of a location or a scale parameter of an arbitrary distribution.     

Efficient sequential estimation for compound poisson processes

This paper discusses uniformly minimum variance, unbiased sequential estimation of a real-valued parametric function for a compound Poisson process when the compounding random variables belong to the exponential family. The characterization of Cramer-Rao efficient plans by Stefanov (1982 b) is shown to be incomplete by obtaining a new efficient plan for the compound Poisson-Bernoulli process. This new plan completes the characterization of Cramer-Rao efficient plans. The class of Bhattacharyya efficient estimators of order two is determined for all the efficient sampling schemes.     

Sequential estimation for continuous time finite markov processes

Efficient sequential estimation of the intensity rates of a continuous-time finite Markov process is discussed. An information inequality which gives a lower bound for the variance of an unbiased estimator of a function of the intensity rates is obtained and it is used to define an efficient estimator. All closed efficient sequential sampling schemes are characterized.     

One way anova for dependent observations

The necessary and sufficient conditions for the classical one way ANOVA procedure to be valid for testing the equality of means of k indepent, first order stationary Gaussian sequences are obtained. Some special models are examined in the light of these necessary and sufficient conditions.