1. This exercise involves doing a content analysis of dating advertisements. Your raw materials will be a page or two of such advertisements taken from a newspaper or other publication containing such advertisements. The focus of the analysis will be on documenting the way in which gender and sexual preference are constructed in the advertisements.

(a) Defining categories: Look through the advertisements and discuss them with others with whom you are doing this analysis. Develop a list of categories and keywords which describe the main attributes that people seek for in partners (for example, words that relate to physical appearance, to character, to social status, to expectations of the desired relationship).

(b) Assigning categories: Once your list is complete, go through each advertisement indicating whether each category applies to each one. Where there are disagreements over the assignment of categories, discuss these. Keep a tally of how many advertisements are assigned to each category.

(c) Analysis: Look at the overall distribution of advertisements across the various categories and try to draw some conclusions about the attributes sought for in partners. Which attributes predominate? How do the attributes vary according to whether men, women, heterosexual or homosexual partners are sought for? Why do you think this is? Have your categories worked well? What different story would other categories have told? Do you think your findings can be generalised to other magazines or media that contain such personal advertisements?

(d) You can apply this process to a variety of other media and to topics other than personal ads. It is often illuminating to make statistical comparisons of different media, or to compare a medium in the past with the same genre today.

(a) Download some electronic text that you are interested in, and divide it into two groups (A and B) sharing common characteristics (e.g., text produced by men and women; articles appearing in 'serious' and 'popular' newspapers). Try to ensure that you have at least 50,000 words in each group.

(b) Use Wordsmith Tools to produce a word list of each of the two groups of text, and save them.

(c) Use the software to produce keyword lists, firstly of A compared with B, then of B compared with A. Inspect the top 100 keywords in these lists to see if you can group words into categories sharing similar characteristics. This will require investigation of some of the words whose predominant meaning in the relevant texts may be unclear, or ambiguous. KWIC displays will be useful here.

(d) For each significant group of words, divide the words into those that are 'key' in group A texts, and those that are 'key' in group B texts.

(e) What has this told you about the characteristics of these texts? Has it revealed anything you would not have known already, or by a brief reading of a few of the texts? What further investigations might you conduct to examine the context in which particular words are being used?