Psychic Distance Stimuli (Dow & Karunaratna, 2006)
Psychic distance stimuli (Dow & Karunaratna, 2006)
The psychic distance stimuli were developed by Dow and Karunaratna for their study originally presented at the Academy of International Business meeting of 2004 and subsequently published in the Journal of International Business Studies in 2006.
While the most widespread definition of psychic distance is “factors preventing or disturbing the flows of information between firm and market. Examples of such factors are differences in language, culture, political systems, level of education, level of industrial development, etc.” (Johanson & Wiedersheim-Paul, 1975), other researchers (e.g. Evans et al, 2000; Sousa & Bradley, 2006; Em, 2019) have argued that it is more accurate to consider that psychic distance has to do with managerial perceptions. Dow credibly argues that the psychic distance stimuli are drivers of those managerial perceptions. Therefore, these measures are not direct measures of psychic distance, but can rather be considered as some of its antecedents.
The psychic distance stimuli are measures of differences between a home and a host country in terms of languages, religions, industrial development, levels of education, and political systems. If you wish to use these scores, I strongly recommend you to check those links so that you have a precise idea of how each score has been calculated.
Professor Dow updated the original data to include more recent years. In his own words: “In 2020, I have revised the distance scales in four ways. (1) Most critically I have updated the scales to include scores for 1995, 2005 and 2015. As part of this process, I have mildly revised the items for measuring industrial development. In most instances the metrics here are very slow to change; thus, 10 years intervals are sufficient. Although researchers can interpolate between these intervals.
(2) Second, I have expanded the set to cover 150 countries, and mildly revised the selection of countries to more accurately reflect the major economies. However, with the latest iteration, I have structured it so that additional countries can be added. Thus, if you desperately need coverage of an additional country, please contact me directly. (3) Third, I have finally given in to requests to create a formative factor that represents the combination of all five of my dimensions. I have used the Mahalanobis distance approach as recommended by Berry, Guillén & Zhou (JIBS, 2010). (4) Fourth, I have re-scaled the distance factors so that they range from 0 to 10, with 0 representing ‘no difference’ and 10 representing the maximum difference. In previous versions, I had left the factors as z-scores, and that resulted in some confusion. Hopefully this re-scaling will resolve that problem.”
The main differences between the original data on Prof. Dow’s website and the data you can download here are as follows: (1) the variables have been renamed to make them more readily understandable (personal preference), (2) a column “year” has been added and the data has been rearranged in order to respect the principles of tidy data (one column = one variable, one observation = one row), and (3) the names of the countries have been replaced by their ISO3 code to increase the ability to merge these datasets with others.
Data on the five psychic distance stimuli for 22 350 country pairs (150 home countries, 150 host countries). The dataset contains 67050 observations across 30 variables:
country_pair: character string uniting the origin and the destination.
origin: character string indicating the first country (ISO3 code) of the country pair.
destination: character string indicating the second country (ISO3 code) of the country pair.
ind_GDPpercap_diff: number indicating the difference between the home (origin) and the host (destination) country in USD GDP per capita (variable name in the original dataset: \(I_1ij\)).
ind_energy_cons_diff: number indicating the difference between the home (origin) and the host (destination) country in energy consumption (equivalent kg coal per capita) (variable name in the original dataset: \(I_2ij\)).
ind_nonagr_labour_diff: number indicating the difference between the home (origin) and the host (destination) country in the percentage of non-agricultural labour (variable name in the original dataset: \(I_4ij\)).
ind_urban_pop_diff: number indicating the difference between the home (origin) and the host (destination) country in the percentage of urban population (variable name in the original dataset: \(I_6ij\)).
ind_phone_diff: number indicating the difference between the home (origin) and the host (destination) country in the number of telephones (fixed and mobile) per 100 people (variable name in the original dataset: \(I_9ij\)).
ind_internet_diff: number indicating the difference between the home (origin) and the host (destination) country in the proportion of the percentage of the population who use the internet (variable name in the original dataset: \(I_{11ij}\)).
PDS_INDUSTRIAL_DEVELOPMENT: number corresponding to the single-factor solution, using principal component analysis of the six previous variables, rescaled so that ‘no differences’ = 0, and ‘most differences’ = 10 (variable name in the original dataset: \(Ind Dev\)).
educ_literate_adults_diff: number indicating the difference between the home (origin) and the host (destination) country in the percentage of literate adults (variable name in the original dataset: \(E_1ij\)).
educ_secondary_educ_diff: number indicating the difference between the home (origin) and the host (destination) country in the proportion of the population enrolled in second-level education – adjusting for the age profile of the country (variable name in the original dataset: \(E_2ij\)).
educ_tertiary_educ_diff: number indicating the difference between the home (origin) and the host (destination) country in the proportion of population enrolled in third level education – adjusting for the age profile of the country (variable name in the original dataset: \(E_3ij\)).
PDS_LEVELS_OF_EDUCATION: number corresponding to the single-factor solution, using principal component analysis of the three previous variables, rescaled so that ‘no differences’ = 0, and ‘most differences’ = 10 (variable name in the original dataset: \(Edu Dist\)).
polsys_political_constraints: number indicating the difference between the home (origin) and the host (destination) country using the POLCON V political constraints scale. (variable name in the original dataset: \(D_1ij\)).
polsys_democ_autocracy: number indicating the difference between the home (origin) and the host (destination) country using the Modified POLITY IV democracy-autocracy scale (variable name in the original dataset: \(D_2ij\)).
polsys_fh_pol_rights: number indicating the difference between the home (origin) and the host (destination) country using the Freedom House Political Rights scale (variable name in the original dataset: \(D_3ij\)).
polsys_fh_civil_lib: number indicating the difference between the home (origin) and the host (destination) country using the Freedom House Civil Liberties scale (variable name in the original dataset: \(D_4ij\)).
PDS_POLITICAL_SYSTEMS: number corresponding to the single-factor solution, using principal component analysis of the four previous variables, rescaled so that ‘no differences’ = 0, and ‘most differences’ = 10 (variable name in the original dataset: \(Dem Dist\)).
lang_dist: a 5-point scale which quantifies the difference between the dominant languages of the home (origin) country and of the host (destination) country (variable name in the original dataset: \(L_1\)).
lang_home_in_host: number based on a 5-point scale based on the incidence of the home (origin) country’s dominant language(s) in the host (destination) country (variable name in the original dataset: \(L_2\)).
lang_host_in_home: number based on a 5-point scale based on the incidence of the host (destination) country’s dominant religion(s) in the home (origin) country (variable name in the original dataset: \(L_3\)).
PDS_LANGUAGES: number corresponding to the single-factor solution, using principal component analysis of the three previous variables, rescaled so that ‘no differences’ = 0, and ‘most differences’ = 10 (variable name in the original dataset: \(Lang Dist\)).
relig_dist: number based on a 5-point scale which quantifies the difference between the dominant religions of any two countries (variable name in the original dataset: \(R_1\)).
relig_home_in_host: number based on a 5-point scale based on the incidence of the home (origin) country’s dominant religion(s) in the host (destination) country (variable name in the original dataset: \(R_2\)).
relig_host_in_home: number based on a 5-point scale based on the incidence of the host (destination) country’s dominant religion(s) in the home (origin) country (variable name in the original dataset: \(R_3\)).
PDS_RELIGIONS: number corresponding to the single-factor solution, using principal component analysis of the three previous variables, rescaled so that ‘no differences’ = 0, and ‘most differences’ = 10 (variable name in the original dataset: \(Relig Dist\)).
psydist_mahalanobis: number indicating a score based on the Mahalanobis Distance approach to combine the five distance dimensions – \(Ind Dist\), \(Edu Dist\), \(Dem Dist\), \(Lang Dist\) and \(Relig Dist\) – into a single index. This approach controls for correlations amongst the underlying dimensions. The resulting factor was rescaled so that ‘no differences’ = 0, and ‘most differences’ = 10 (variable name in the original dataset: \(Psy Dist_{Mahal}\)).
Reference: Dow, D. and Karunaratna, A. (2006): Developing a Multidimensional Instrument to Measure Psychic Distance Stimuli. Journal of International Business Studies, 37, No. 5, pp. 575 – 577