Making the Randomized Response Technique Work

The ability of the randomized response technique to obtain sensitiveinformation on surveys was compared to two other methods involvingself-administered questionnaires. Using a national probabilitysample, and extending upon a number of developmental changeswhich have evolved over the years, the randomized response techniqueproved to be decidedly superior to the other methods.     

The impact of anti-Vietnam demonstrations upon national public opinion and military indicators

In an attempt to assess the national impact of anti-war protest, 15 major anti-Vietnam demonstrations occurring in the United States from 1965 through 1971 were analyzed for their effects upon Gallup public opinion survey data, American troop strength, and American munitions expenditures. While weak and short-term counterproductive effects appeared for two opinion indicators, an equally likely conclusion is that the demonstrations had little or no lasting impact whatsoever. The possible reasons for this finding, and its social implications, are discussed.     

Product Introduction Strategies under Sequential Innovation for Durable Goods with Network Effects

When network effects are important and technology is rapidly improved, this study explores the relative optimality of five product introduction strategies of a durable goods manufacturer: (1) replacement, (2) skipping, (3) a delayed line, (4) shelving, and (5) line‐extension. Using a two‐period analytical model, we show how the type of compatibility—either full or backward compatibility—and the magnitude of the network effect influence the manufacturer's preference for the above strategies. Our analysis reveals that only the strategies (1)–(3) above can be optimal; and the optimal strategy varies with network strength. Further, the type of compatibility can dramatically change the profitability under each optimal strategy; for instance, while backward compatibility can increase the profitability of replacement under certain conditions, it always reduces the profitability of a delayed line. We also illustrate that if compatibility were a choice, although backward compatibility may be observed widely in practice, the parametric region for its optimality is relatively more restricted than that of full compatibility.