Data collection differences could be an explanation, but there could be others.
Before I retired, I sometimes worked with the Federal Trade Commission on merger cases. To help analyze whether a proposed merger will harm competition and violate the antitrust laws, you need to analyze the geographic market for the products made by the merging parties. In that analysis, the FTC economists take into account the presence of state lines and physical barriers such as rivers, because they have found over the years that people's shopping habits are shaped by those factors in a manner that is statistically significant. That is, on the whole people are reluctant to cross state lines and rivers to shop or eat dinner. Obviously not everyone, but enough to affect the data. When I first learned that, I was surprised, but then I looked at my own habits. I live in a town on a river, and I find that if I leave my own town, it is almost always to go to other towns on my side of the river, even though they are farther away than the town on the other shore and even though there are multiple bridges.
In the case of Texas and Oklahoma, you have not only the state line, but also a physical barrier, the Red River, between them. On average, people in Oklahoma are more likely to stay in Oklahoma as they go about their daily business, and people in Texas are going to stay there. In light of that, I believe there very well could be differences in the prevalence of Covid that are NOT due to differing data collection methodologies.