Jeremy Smith and Michael McAleer, "Alternative Procedures for Converting
Qualitative Response Data to Quantitative Expectations", Journal of
Applied Econometrics, Vol. 10, No. 2, 1995, pp. 165-185. 

The file sm-data.zip contains 9 separate files within it.

(1) manufq.sm contains data on the original variables output, employment,
prices, stocks of unfinished goods and overtime for the manufacturing sector
in Australia.

(2) last.sm contains data on the percentage of firms reporting a rise,
stay-the-same and fall in each of these variables (output,
employment, prices, stocks, and overtime) in the last three months.

(3) forward.sm contains data on the percentage of firms reporting a rise,
stay-the-same or fall and in each of the variables output, employment,
prices, stocks, overtime in the next three months.

(4) doutf.sm contains data on the percentage change in the variable for output
as well as the expectations formed using AR(2), probability, Anderson, Pesaran,
TVP and dynamic non-linear methods.

(5) dempf.sm contains data on the percentage change in employment and the
expectations using AR(2), probability, Anderson, Pesaran, TVP and dynamic non-
linear methods.

(6) dpf.sm contains data on the percentage change in prices and the
expectations using AR(2), probability, Anderson, Pesaran, TVP and Dynamic non-
linear methods.

(7) dstkf.sm contains data on the percentage change in stocks and the
expectations using AR(2), probability, Anderson, TVP and dynamic non-linear
methods.

(8) dotf.sm contain the data on the percentage change in overtime and the
expectation using AR(2), probability, Anderson, Pesaran and dynamic non-linear
methods.

(9) Orthog.sm contain the variables exports, imports, terms of trade (tot),
price of inputs (Pinp), Number of Vacancies (Vacan) and the number of
industrial disputes (Inddis) used in the orthogonal regressions.