Appendix A. Appendix A: Adjusted Median Salary

GEOGRAPHY AND EXPERIENCE CLEARLY MAKE A DIFFERENCE IN SALARY, and this is fully expected. However, since geography and experience can correlate with other variables, unless we analyze all three variables together, it can be hard to tell whether variations in salary are due to these variables or to geography/experience.

For example, age correlates with experience and with salary, but if we consider groups of respondents with equal experience, then age no longer correlates with salary (at least, strongly or monotonically). This is what we mean when we say that age and salary don’t correlate when we “block” years of experience.

To give another example, this time with geography: the median salary of the 9% of respondents who say they code over 20 hours/week is $50K, while the rest of the sample (those who spend less time coding, if any at all) is $80K. However, the difference is attributable to the fact that most of the people who code over 20 hours/week happen to come from places that have lower salaries in general. For example, 36% of respondents from India, Russia, and Ukraine say that they code over 20 hours/week, while only 1% of California respondents do. This probably shouldn’t be taken to mean that CA design professionals don’t code: correlations like this will appear frequently on such surveys; namely, when there is little control over the sampling. This correlation is likely just noise that we should try to filter out.

The solution ...

Get 2017 Design Salary Survey now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.