Nonetheless, it has been illustrated that the standard font of a typeface is not the best per sé for every application. In a comparison of Memphis fonts of different degrees of boldness, Luckiesh and Moss (1940) found Memphis Medium to be of higher readability than the other fonts when all were 10 points in size and set with 2 points of leading. Memphis Bold, however, was the most visible of all. The term visibility here seems to correspond to the concept of conspicuity. It was concluded that using bolder versions of many typefaces than those versions used for body text in the 1930’s would result in higher readability. Body text is the main, regular text, to be distinguished from consultation text. Consultation refers to units of text that provide additional information in, for example, margins and boxes. Another conclusion of the study was that the "optimum degree of boldness is much more sharply defined by the criterion of readability than by that of visibility".
Like italics and case, boldness is often given a specific role in text; words are often emphasized by making them bolder than the surrounding text. Within a sentence, bold words tend to draw too much attention (McAteer, 1992). Therefore, boldness is generally not given the function for which italics or upper case are used, but is rather used for headings and labels.
When a typeface has been chosen, based on the considerations offered above, a number of other text characteristics still need to be determined. Søgren (1995) proposes an order in which these characteristics need to be given attention, suggesting that it is a typographic convention. He states that choice of typeface is to be followed by choice of typesize, after which letter spacing, word spacing, justification, line width, and leading need to be determined, in that particular order. Naturally, all this is done only after the dimensions and margins of the printable surface have been determined. Although common sense seems to agree with Søgren, no formal support or disagreement can be found for the proposed order. Perhaps some designers may disagree with the implication that design is a serial process, but one would think that it is excusable for the purpose of evaluation. Since no alternative order with more obvious merit has been suggested, all steps in the typographic arrangement of text will be discussed here and in chapter 3 according to this order. Freedom of choice in every step is of course affected by the choices that have been made in previous steps. Therefore, effects of interaction between aspects of typography and the statistical significance of such interactions will be discussed where research is available.