As noted in Table 8, House et al. (2004) suggest nine cultural construct values in comparison to Hofstede's five cultural values. They designed their questions based on known cultural values that were theoretically derive; hence removing the first criticism for Hofstede's values: lack of theoretical basis when he developed them. Given that House et al. collected the data between 1994 and 1997, the cultural values reported in GLOBE are rather new and only about 10 years old; hence addressing the issue of outdated data.
It is important to mention that GLOBE has also been subject to criticism (e.g Hofstede, 2006). Hofstede mentions it as being too abstract. Smith (2006) mentions that the aggregation of individual survey responses to the cultural level of analysis is problematic. Earley (2006) suggests that large-scale multi-country studies of managerial beliefs may not measure organizational phenomena and actual behavior. On the other hand, many researchers have found the new index more comprehensive than the ones provided and used in prior research. Based on the definition of the GLOBE and Hofstede's values, we tried to make a comparative table between the two cultural indexes and run similar tests.
Note that the new tests use a different and more current data set (earnings management data is calculated for 1994 to 2005). Furthermore, the number of observation and countries is less (52,857 observations in 25 countries; Belgium and Norway were missing in GLOBE). So, our results in sensitivity analysis using GLOBE may result in a different outcome because of the change in database. To alleviate this difficulty, we tested both Hofstede and GLOBE values using the new data sets. our results are mostly similar to those reported for the original (and older data). Hence we are able to provide further evidence for our suggestions in this study. Only in the