¹ Studies of each type described in this paragraph are cited in Scarritt, James R., Dimensions of Change in the Zambian One-Party Second Republic (forthcoming), chap. 1.Google Scholar

² This distinclion is made with regard to human rights measurement in Stohl, Michael et al. , ‘State Violation of Human Rights: Issues and Problems of Measurement’, Human Rights Quarterly, 8 (1986), 592–606, p. 606.CrossRefGoogle Scholar

³ Hermann, Charles F. et al. , CREON: A Foreign Events Data Set (Beverly Hills: Sage Professional Papers in International Studies, 2–024, 1973), p. 19.Google Scholar

⁴ Africa Contemporary Record (New York and London); Africa Research Bulletin: Political Series (Exeter); and African Recorder (New Delhi).

⁵ On the first point, see Burrowes, Robert, ‘Mirror, Mirror on the Wall…: A Comparison of Event Data Sources’, in Rosenau, James N., ed., Comparing Foreign Policies (New York: John Wiley and Sons, 1974), pp. 383–406Google Scholar; Hoggard, Gary D., ‘Differential Source Coverage in Foreign Policy Analysis’, in Rosenau, James N., ed., Linkage Politics (New York: The Free Press, 1974), pp. 353–82Google Scholar; Jackman, Robert W. and Boyd, William A., ‘Multiple Sources in the Collection of Data on Political Conflict’, American Journal of Political Science, 23 (1979), 434–57CrossRefGoogle Scholar; and Stohl, et al. , ‘State Violation of Human Rights’, p. 604.Google Scholar On the second point, see Ben-Dak, Joseph D., ‘Some Directions for Research Toward Peaceful Arab–Israeli Relations’, Journal of Conflict Resolution, 16 (1972), 281–95, p. 291CrossRefGoogle Scholar; Bollen, Kenneth A., ‘Political Rights and Political Liberties in Nations: An Evaluation of Human Rights Measures, 1950 to 1984’, Human Rights Quarterly, 8 (1986), 567–91, p. 590CrossRefGoogle Scholar; and Shaw, Timothy M. and Anglin, Douglas G., Alternative Sources of Event Data on Zambian Foreign Policy (Syracuse, NY: Maxwell School, Syracuse University Foreign and Comparative Studies/African Series, 36, 1981).Google Scholar

⁶ The Times of Zambia was owned by the governing United National Independence Party (UNIP) between 1975 and the end of the one-party system, but retained considerable independence in the determination of news coverage. See United States Department of State, Country Reports on Human Rights Practices for 1985 (Washington: US Government Printing Office, 1986), p. 384Google Scholar; Andreassen, Bard-Anders, Geisler, Gisela and Totensen, Arne, Setting a Standard for Africa? Lessons from the 1991 Zambian Elections (Bergen: Chr. Michelsen Institute, 1992), p. 41.Google Scholar; Kasoma, Francis P., The Press in Zambia: The Development, Role and Control of National Newspapers in Zambia 1906–1983 (Lusaka: Multimedia Publications, 1986), pp. 82–108, 139–59Google Scholar; and Lungu, Gatian F., ‘The Church, Labour and the Press in Zambia: The Role of Critical Observers in a One-Party State’, African Affairs, 85 (1986), 385–410, pp. 390–4, 405–9.CrossRefGoogle Scholar The coding of policies in the larger study of which this Note is a part included government actions towards the newspapers, and these reveal that when editors were fired, those appointed to replace them soon took equally independent positions.

⁷ See Bollen, ‘Political Rights and Political Liberties in Nations’, pp. 582–3; Gurr, Ted Robert, Politimetrics: An Introduction to Quantitative Macropolitics (Englewood Cliffs, NJ: Prentice-Hall, 1972), pp. 83–4Google Scholar; Hopple, Gerald W. and Kuhlman, James A., ‘Expert-Generated Data: An Overview’, in Hopple, and Kuhlman, , eds, Expert-Generated Data: Applications in International Affairs (Boulder, Colo.: Westview Press, 1981), pp. 19–27Google Scholar; McCamant, John F., ‘A Critique of Present Measures of “Human Rights Development” and an Alternative’, in Nanda, Ved P., Scarritt, James R. and Shepherd, George W. Jr, eds, Global Human Rights: Public Policies, Comparative Measures, and NGO Strategies (Boulder, Colo.: Westview Press, 1981), pp. 129–30Google Scholar; and Fred R. von der Mehden and Kim Quaile Hill, ‘Methodological Issues and Problems in Employing Expert-Generated Data’, in Hopple and Kuhlman, Expert-Generated Data, pp. 39–43.

⁸ The latter were identified by an informant who has made a career of studying the press in Zambia.

⁹ Panel size was limited by the number of genuine experts whom it was anticipated would be willing to participate, but a large panel was sought both because of the great variety of information contained in the events and because of the desire for political, occupational and national balance. This panel is larger than recommended by Von der Mehden and Hill. ‘Methodological Issues and Problems in Employing Expert-Generated Data’, pp. 44–5; it is smaller than that used by Russell H. Fitzgibbon and Kenneth F. Johnson. See Johnson, , ‘Scholarly Images of Latin American Political Democracy in 1975’, Latin American Research Review, 11 (1976), 129–40Google Scholar, and citations to earlier versions of the same study contained therein.

¹⁰ Careful consideration was initially given to Guttman scaling, but this proved to be too exacting a test for the complex and subjective evaluations that judges in this project were asked to make, in terms of both its relatively rigid parametric assumptions and its consequent inability to accept any more than minimal deviations from perfect scales. See Mclver, John P. and Carmines, Edward G., Unidimensional Scaling (Beverly Hills, Calif.: Sage Quantitative Applications in the Social Sciences, 1981), pp. 40–71CrossRefGoogle Scholar; and Mokken, R. J., A Theory and Procedure of Scale Analysis (The Hague: Mouton, 1971), pp. 23–71.CrossRefGoogle Scholar In addition, this method cannot effectively handle the very large number of items (events) presented for scaling in this project.

¹¹ Mokken, A Theory and Procedure of Scale Analysis, describes the technique on pp. 170–99 and several examples of its use in political research on pp. 200–332. A good summary of the technique is presented in Kingma, Johannes and Taerum, Terry. ‘SPSS-X Procedure and Standalone Programs for the Mokken Scale Analysis: A Nonparametric Item Response Theory Model’, Educational and Psychological Measurement, 49 (1989), 101–36, pp. 103–21.CrossRefGoogle Scholar More recent applications in political research than those cited by Mokken, which illustrate the variety of topics on which the technique has been utilized, include Gillespie, M., TenVergert, Elisabeth, and Kingma, Johannes, ‘Using Mokken Methods to Develop Robust Cross-national Scales: Do the Six Abortion Items in the NORC GSS Form One or Two Scales?’ Quality and Quantity, 21 (1987), 393–408Google Scholar; and Stokman, Franz N., Roll Calls and Sponsorships: A Methodological Analysis of Third World Group Formation in the United Nations (Leiden: Sijthoff, 1977).Google Scholar

¹² Since a significant minority of judges (the exact number varying among categories of events) found it difficult to perform the additional task of ranking major events, this dichotomization did not involve much loss of information.

¹³ In terms of the present project these assumptions can be stated as: (1) the probability that a judge will select a given event as major increases with his or her propensity to select events as major, and (2) the probability that a given event will be selected as major increases with the number of judges who select it as major or, in other words, the ordering of events remains constant over judges. Coefficients of scalability between each pair of events (H_ij), between a specific event and all other events in a potential scale (H_i), and for the scale as a whole (H) are used, in conjunction with the significance test Δ*, to test the first of these assumptions. In addition to a significant Δ*, all H_ijS must be greater than zero and all H_jS (and therefore H) must be greater than some constant, which Mokken specifies at 0.5 for a strong scale, the level that is utilized in the present research. To test the second assumption, the probability matrices for positive and negative responses (designating events as major or not so designating them) for each judge are inspected to discover how close each comes to the response predicted by her or his scale score, and the statistical significance of departures from the expected pattern of responses is tested by a runs test. Events not meeting both of these assumptions are eliminated in order to form one or more scales from the remaining events. Since these scales are non-parametric, deviations from perfect scales – in which each judge selects as major all events selected by lower scoring judges plus one or more additional events – are permissible if they fall within the limits of statistical significance established by the researcher.

¹⁴ Kingma and Taerum, ‘SPSS-X Procedure and Standalone Programs for the Mokken Scale Analysis’; and TenVergert, Elisabeth, Kingma, Johannes and Taerum, Terry, ‘Psychology of Computer Use: VIII. Utilizing a Nonparametric Item Response Model to Develop Unidimensional Scales: Mokscal’, Perceptual and Motor Skills, 68 (1989), 987–1000.CrossRefGoogle Scholar

¹⁵ Mokken, A Theory and Procedure of Scale Analysis, p. 196.

¹⁶ See Conge, Patrick J., ‘The Concept of Political Participation: Toward a Definition’, Comparative Politics, 20 (1988), 241–9CrossRefGoogle Scholar; Huntington, Samuel P. and Nelson, Joan M., No Easy Choice: Political Participation in Developing Countries (Cambridge, Mass.: Harvard University Press, 1976), pp. 12–15CrossRefGoogle Scholar; and Verba, Sidney, Nie, Norman H. and Kim, Jae-on, Participation and Political Equality: A Seven Nation Comparison (Cambridge: Cambridge University Press, 1978), pp. 51–6.Google Scholar

¹⁷ A partial exception is a brief statement found in Bratton, Michael, ‘Zambia Starts Over’, Journal of Democracy, 3 (1992), 81–94, p. 84.CrossRefGoogle Scholar

¹⁸ Gertzel, Cherry ed., with Baylies, Carolyn and Szeftel, Morris, The Dynamics of the One-Party State in Zambia (Manchester: Manchester University Press, 1984).Google Scholar

¹⁹ The first two digits in the event numbers listed in this article are the year in which the event occurred; the second two digits are the event number within that year; and the last two are the category into which the event was placed as part of its coding. Category 04 is changes in voting behaviour; category 05 is entering or withdrawing from political roles; category 06 is changes in party organization or party–government relations; and category 15 is changes in policies toward participation, (NS indicates that the event is not part of a scale.)

²⁰ Mokken assigns events the scale position that maximizes their fit into a scale, and this may differ slightly from their actual frequency of selection.

²¹ Baylies, Carolyn and Szeftel, Morris, ‘The Fall and Rise of Multi-Party Politics in Zambia’, Review of African Political Economy, 54 (1992), 75–91, p. 77.CrossRefGoogle Scholar