Both natural and social sciences statistically analyze empirical observations about their target domain: e.g., physics analyzes observations of waves, particles, atoms, materials, etc. while economics analyzes observations of prices, wages, job openings, unemployment, etc.
Each special science (whether natural or social) can also differ from the others in terms of its tools for data collection (e.g., particle colliders in physics, spectrophotometers in biology, and electroencephalographs in psychology).
The humanities are also interested in empirical observations, but their analysis is logical, historicall, literary, aesthetic, etc. For example, philosophers of science are often trained in science and have even done some science themselves, giving them familiarity with the statistical analyses of empirical observations. However, philosophers are often analyzing the logic of the arguments employed by scientists’ (often unnoticed) philosophical claims—e.g., about how data should be collected and analyzed or what conclusions should be drawn from data. These arguments often involve assumptions that are not supported (or else unsupportable). For example, philosophers have often pointed out the problem of induction: that data can only be compatible with a hypothesis rather than “support” a hypothesis.
So there is more similarity than difference between the sciences and humanities: both are (or should be) firmly interested in empirical observations. However, each special science and subset of the humanities uses slightly different tools and analytical frameworks—not because of differences in their rigor, but because of differences in their goals (e.g., statistical predictions vs. logical evaluations) and the targets of investigation (e.g., atoms vs. cells vs. minds vs. scientific arguments).
