It is probably evident that all sources are not created equal. Some are much more equal than others. (Attribution to George Orwell). As I have been doing research, I have noticed that source records are often buried in layers deposited over time. The layering effect comes as a result of the time of deposition of the record, the nature of the record itself, the availability of the record and many other factors. Unfortunately, many genealogical researchers are not aware of the existence of these layered records and when they reach the lowest level in one layer, they think their research is through even though there are many, many more layers of records left to research.
Now I think I need to give some examples of what I am talking about. Note, that any effort I make to assign a particular type of record or source to a "layer" is completely arbitrary. The existence of layers of records is my personal way of viewing the records and not any actual distinction between different types of records. But I think this way of looking at records and sources helps to keep me searching when I know that another layer of records may be uncovered at any time. In my way of thinking, records and sources in lower layers are more difficult to find and research than those in high layers.
The top layer: of records are those that are easily obtainable and on their face refer to other possible records. The prime example of the top layer of records is the U.S. Census (or any other national census record). These records are well known and with all the electronic digitized copies online, easy to search. The type of information given by Census records is usually helpful in identifying families and family members. Finding a census record of a family suggests, at the very least, marriage records, birth records, land records, employment records and many other types of records. Other records that fall in the top layer are those such as the Social Security Death Index and other records that record specific information about individuals and families over a wide geographic area.
You move down to the next layer when you begin to look for records that have a more limited geographic coverage. The reason why these records are different than the Census records is that you must know where your individual ancestor or family lived before you can locate the record. Vital Records fall in this second category or layer, as do land records and many other records of a more restricted geographic application. One reason why these records are substantially different than top layer records such as the U.S. Census, is that at this layer, many of the records are not online and/or readily available.
Any time you move back in time over 50 or so years, you also move down a layer. For example as you move back in the past, vital records such as birth, marriage and death certificates start to disappear, they only live in the topmost layers of records. So you may be at the top layer of records but as you move back into the past, you find the records more difficult to research or even to find. A current issue is also online availability. If a record is not available online, it automatically moves into a lower layer of research difficulty.
The next lower layer of records is composed of records that become harder to research because of their nature. Military records belong to this layer and of course, as you go back in time, military records are less and less available. Some church records also belong in this layer. Not because they aren't valuable, but because they require specialized research techniques.
By the time you get to the next layers, you have exhausted all of the readily available online records. In this next lower level of availability reside records such as tax records, cattle brands, water records and other such records that only have a dim existence in the normal, light-of-day genealogical world. Not only are these records obscure, they are also fairly hard to find.
At the bottommost layers lie the really valuable gold-bearing records. These are the ones that have been so seldom examined that there are few references to their existence. I include in these records the contents of many historical societies and university and college libraries. Not only are these records hard to find, they also may reside at a considerable distance from the source. Although they may be cataloged by the institution, the catalog entries likely tell the researcher little or nothing about the detail of the content. But these types of records are like gold. If you find the gold, you get rich. If you don't find the gold, you go broke.
The interesting thing is that many researchers are unaware that there are any lower layers of records. When they don't find what they are looking for in some superficial layer, they do not know enough to dig deeper and go down a separate layer.
Good luck searching the layers.