These are points you allocate in order to gauge the ratings and personality of a prospect. You have a fixed amount of scouting points each week, representing the resources you have available to travel the country and look at prospects. After each sim you get all the scouting points back, and can allocate them again.
The total amount of points you can spend scouting a player is 20. The amount of points you allocate stack meaning that you dont have to spend 20 at once, but can pour in a little each sim. The more points spend on a player, the more accurate the assessment of his ratings will be. This means that the assessment of his ratings in a case where you have only assigned one point will be significantly less reliable than a case where you have spent 15. When you reach a total of 20 points spent, you will have the most accurate report you can get. You can track the reliability on the 'report accuracy' bar on the player card.
Understanding the ratings assessments
When you look at a player you haven't scouted yet, all his ratings will be listed as unscouted. When scouting a player, your scout will come back with an assessment of each of his relevant ratings, this means your scout wont be bothered with assessing the tackling of a QB and so on. The most important part to understand about the assessments is that they are based on the average of that particular rating across all active players of the same position. This means that if you are scouting a TE and getting an assessment of his Hands skill, the scout will base his assessment on the average of the Hands of all TE's currently in the league. If the average of all TE's Hands are 63, and the player you are scouting has 63 hands the scout will tell you his hands are average (if he is 100% accurate). Now, because there is a lot of really bad players on really bad teams this means that those averages may be lower than you expect. If you are a top tier team, then 'average' Hands might not cut it for you if you want a pass catching TE.
Margin of error in the assessments
There are 22 (not counting 'unscouted') different assessments ordered by their relation to the benchmark. Observe that the margin of error in your scouting report will show up in how far removed the given assessment is from the correct assessment in the ordering. This means that as you increase your report certainty the error margin will decrease, and you may be satisfied with less than full certainty when you feel the given assessment likely indicates the approximate level (i.e. the error margin is reduced and you like the indication it is giving).
List of ratings
Remember, these are dynamic and the benchmark will change depending on the player pool, so not only will it differ between positions, it will also differ from season to season as the player pool changes. Additionally, because of the number of different ratings, only a few points separate each.
Interviews are crucial because they unlock two things. First and foremost, they allow you to see a players personality, which is important in order to get an idea of what he likes, and assess your chances of signing. Secondly, interviewing a player reveals whether he has any special traits (listed under 'special' on his player card). Traits give various bonuses either on or off the field, and thus make the player more valuable. You can only interview a player once, but on the positive side, there is no error margin on the information you get. Scheduling an interview costs 5 scouting points.