A few years ago our organization decided to have a panel discussion with students that we teach in online courses. The information that I took away from the panel was that students needed access to recorded classes (they thought it would be nice to have them in mobile format), needed access to sample material for assignments/tests and wanted short videos for extra help and remediation. 
Yesterday I was having a conversation with a fellow colleague about our online students. More specifically it was about what they access asynchronously, and what they don't. How does this tie into what I took away from the panel discussion? Let me use some of the student analytics taken from Desire2Learn (D2L) to explain.

In the first graphic you can see that there are some students that watch some recordings. However, in a class of 22 students, the number that watch the recordings is small. Furthermore if you look at the average time that students have the recordings open, in addition to the fact that many of these recordings are 40 minutes in length, the conclusion to reach is that even those that open the class do not watch the class from start to finish. The unplugged recordings allow students to open the recording and print notes. As well students can rewind/ fast forward the recording. Are they opening the recording for the notes? Skipping the areas of the recording that they do not need? Either way most students do not access the recordings and those that do may not be watching them from beginning to end.
The mobile piece was surprising to me. If you look at the same graphic above you will notice that despite the same classes being available in mobile format, the recordings in this format were rarely, if ever accessed. I know students are not accessing them but the question is why. Is it difficult to navigate within the D2L mobile user interface? Do they prefer using their computer? Do they only watch recordings in school where mobile devices are often banned? I'm not sure. But their lack of uptake did surprise me.

I also placed inside my D2L content videos by BrightStorm which do a really good job in keeping content short and to the point. Students have had access to these videos throughout the unit we were covering. I even gave class time to watch these. Again, in a class of 22 students, the viewing of these was minimal. Furthermore, these videos are accessible on all devices, iOS included. If the length of the recorded classes were an issue, I would expect students to access the videos that were shorter in length.
For every worksample I give as a part of my formative assessment and for every test that I give as a part of summative assessment, there are sample materials under content in D2L. In the third graphic you can see the use of those sample materials, especially for testing purposes, is not where I would expect it to be. Given the graphic is from an advanced mathematics class, the analytics data is not what I would have expected.
The final graphic shows the analytics from the worksamples and tests I have given my class this year. If you look at the numbers reported for the solutions you will see that many students are not accessing the solutions to worksamples and tests. That means students are not going back over their work to see where they made mistakes. Again, these are math students in an advanced class.
Perhaps the most surprising statistic is their overall result. Students in this class have a mean grade very close to 80%, with a standard deviation of about 9%. The analytics are telling me that students are not accessing what they need to for success, but are achieving it anyways. Why?
First of all the synchronous classes are still the place where content is delivered. As well I answer many questions in class when students have trouble.
Secondly, many of the students I teach are still in a school with a math teacher on site. I know they go to their face to face teacher for help... they tell me.
So what do I conclude from the D2L analytics? I still believe students do want access to information and help on demand. However, I believe that when the on demand component can be synchronous, whether it will be from a classmate, face to face teacher, or even an online teacher/tutor, students will gravitate to that. Asynchronous may not be their first choice if there are synchronous supports there.
My colleague and I feel that students are not accessing content outside the regular school time, or building. The analytics provided by the D2L content manager, to a large extent, seems to confirm this. The analytics seem to contradict what students were actually saying in the panel discussion. One thing is for certain... as learning becomes more blended and components moved to online, analytics that a Learning Management System like Desire2Learn provides is absolutely critical to helping teachers filter through what students say they need... and what they actually do.