Tattooing and Techniques of Cross-Cultural Research
Part II: Derive Hypotheses

by William Divale

 Department of Anthropology

York College (City University of New York), Jamaica, NY

HYPOTHESIS A

Tattooing is a painful process. Societies tend to be consistent, so one would expect that if pain is inflicted in one realm, it will be inflicted in others, such as corporal punishment of children.

Cross-Cultural: Thus, if the above is true, one could expect that societies with tattooing will also tend to have higher corporal punishment of boys and girls.

Table 1. Cross-tabulation of Corporal Punishment of Boys in Early Childhood and Tattooing. 

Notice above that the societies with lower levels of corporal punishment for boys (1-5) seem fairly evenly distributed between having tattooing present and absent. However, the societies with high levels of corporal punishment (6-9) almost always have tattooing. This suggests that there is a trend consistent with the hypothesis. The statistics that test for a linear association are shown in "Symmetric Measures." Since the variables are ordinal, the appropriate measures of association are gamma (.463) and Spearman's correlation (.328). The coefficients of association are moderate and indicate a positive association (the more corporal punishment, the more tattooing is present. The significance levels are .058 and .076 respectively. Although these significance levels are marginally significant, we are entitled to halve the probabilities since the association is in the directed predicted. Thus, the probability that the results are due to chance is .029  (29 chances out of 1000) using gamma and .038 (38 chances out of 1000) using Spearman's rho. Thus with regard to corporal punishment of boys we can reject the null hypothesis that the results are due to sampling error, and we accept the theoretical hypothesis that tattooing and corporal punishment of boys are positively related.

Notice above that the societies with lower levels of corporal punishment for girls (1-5) seem fairly evenly distributed between having tattooing present and absent. However, the societies with high levels of corporal punishment (6-9) almost always have tattooing. This suggests that there is a trend consistent with the hypothesis. The statistics that test for a linear association are shown in "Symmetric Measures." Since the variables are ordinal, the appropriate measures of association are gamma (.506) and Spearman's correlation (.367). The coefficients of association are moderate and indicate a positive association (the more corporal punishment, the more tattooing is present. The significance levels are .037 and .050 respectively. We are entitled to halve the probabilities since the association is in the directed predicted. Thus, the probability that the results are due to chance is .015  (15 chances out of 1000) using gamma and .019 (19 chances out of 1000) using Spearman's rho. Thus with regard to corporal punishment of girls we can reject the null hypothesis that the results are due to sampling error, and we accept the theoretical hypothesis that tattooing and corporal punishment of  girls are positively related.

HYPOTHESIS B

1.Tattoos are associated with less sexual guilt.

Cross-Cultural: If the above is true, one would expect that societies with tattooing would have more sexual expression for adolescent boys.

Sexual Expression in Adolescent boys and girls  - Encouragement of sexual behavior, taking into account its frequency, emotional intensity, importance, and variety (including range in partners) in adolescence.  Heterosexual intercourse is the principal criterion, but heterosexual foreplay, masturbation, homosexuality, sexual jokes and exposing the genitals were also considered

Table 2. Cross-tab Between Tattooing and Adolescent Boys Sexual Expression

The results in Table 2 indicated a very weak association between the presence of tattooing and boys sexual expression.  If one looks at the Tattoo Present column, there does not appear to be any pattern between tattooing and sexual expression.  The measure of the strength of association, Gamma is very low (-.235) and the probability the results are due to random sampling error is .379 or 379 chances out of 100.  Thus we have to accept the null hypothesis that these two variables are not related and the hypothesis is not supported.

 
2.Tattoos are associated with less sexual guilt.

Cross-Cultural: If the above is true, one would expect that societies with tattooing would have more sexual expression for adolescent boys.

Table 3.  T-Tests Between Tattooing and Sexual Non-restraint in Adolescents

Range of Sexual Non-Restraint:  1=Strictly Prohibited to 9=Condoned & not punished

The higher score on sexual non-restraint means that non-restraint is condoned, e.g., more sexual expression is allowed.  For both adolescent boys and girls the scores are higher when tattooing is absent (7.20 for boys and 6.70 for girls).  Sexual expression scores are lower when tattooing is present.  This goes in the opposite direction of the hypothesis.  The differences in means between the two groups is also not significant (.105 for boys and .216 for girls), which means that the differences in means between the tattooed and non-tattooed societies could be the result of chance sampling error.  Thus we have to accept the null hypothesis.

INDEX

Part I: Syllabus
Part II: Outline of Basic Steps of a Cross-Cultural Study
(contains Proposition Inventories, Data Entry Sheet, etc.)
Part III: Outline of a Cross Cultural Study Paper
Part IV: Outline of a Conference Paper