Hall/ Hofstede vs. McLuhan: Sorting out cultural and technological forces impacting U.S. and Korean
political websites
YY
links and how to use interactivities, no cultural difference or entirely shared acoustic space
might happen in the new medium because both interactivity and non linearity are main
origins of expanding space, bringing McLuhanism to the online world.
Limitations and Suggestions for Further Study
Even though this study revealed notably different results from those of previous
studies, it has some limitations to be addressed in further research. Most importantly, a lot of
studies of political advertising and public relations have been conducted during each
campaign period. In fact, persuasive political communication is in its peak during campaigns.
Relatively, political public relations is more noteworthy than for a non-campaign period such
as the one in the current study.
There is another important factor that could influence results. Patriotism or
collectivism strengthened after Sept. 11, 2001, in the United States might be relevant to
collective menus relatively dominant in the U.S. senators’ Web pages. Indeed, lots of national
flags and menus to order them were observed in the U.S. samples and doubtless, national
flags are symbols of collectivism. This study does not prove the effect of an outer factor like
9/11, but such an event must be considered as a qualification of the results.
Some researchers could point out a statistical problem in this study. This study used
the chi-square to analyze the data. In many cases, cells in the tables had frequencies less than
five. There are contradictory arguments for that point.
Siegel and Castellan (1988) noted that …“no more than 20 percent of the cells
should have an expected frequency of less than 5, and no cell should have an
expected frequency less than 1” (p. 199). Some researchers, however, have
found that the chi-square is very robust against a violation of this requirement.
For example, Good, Grover, and Mitchell (1977) suggest that expected values
may be as low as 0.33 without altering the probability of committing a Type I
error (Richard P. Runyon et al, 1999, p. 598).
For example, if there is a medical research to see ovarian cancer incidence in two populations
with one being entirely male, the cell will by definition be zero, but significant differences in
Convention