Tuberculosis and the "social complex" in Glasgow.

Abstract

1 in 20. In this study, however, the more rigorous "F" test* of the ratio of "explained" variance to residual variance is used to judge the level of significance of any relationship expressed as a regression. For significance levels of partial correlation and regression coefficients, the "t" testt is applied. In addition to the tables of data and results of the analyses outlined above, diagrams illustrate the closeness and consistency of the associations found. Scatter diagrams, with regression lines or curves fitted, indicate the progressive increase in disease rates as social conditions in the wards deteriorate, and show the nearness of the actual ward rates to the regression estimates. Variance analysis diagrams, showing the relationships involving each social factor singly and in combination, also demonstrate the small proportion of the variance unaccounted for by the complex relationships.