The short answers: Literature reviews

No, and here's why …

1. Different universities and even different departments have different expectations. For example, some want only articles from refereed journals in the last five years. Others expect a wider range.
2. Some people write an annotated bibliography first and then convert it to a literature review. Others go straight to the literature review.
3. Topics can be put together in different ways. For example, a researcher can tell a story of how thinking changed over time. Or he/she could group things together by theme.
4. Some literature reviews are integrative, when all writers are saying essentially the same thing in different ways. Some are contrastive when views differ markedly.

Short answer is that you can't. You need information (data) to analyse and from which you can draw conclusions.

I know of several cases that had no literature review because nothing had been written on the topic, but the research was ethnographic so they depended on lots of interviewees as sources. It is now very rare to find research topics on which nothing has already been written.

Yes, it is allowed. In fact, you might even find that it is necessary because the literature review should represent the current state of research on the topic. The catch is that you must still give full references, as if it was somebody else’s work. This prevents self-plagiarism, which is the banned practice of recycling your previous work as if it were new. If you self-plagiarize, the penalties in your institution will probably be the same as those for any other kind of plagiarism.

I can't see how students could do too much relevant research, but they often do too much irrelevant research. They easily get interested in topics that are irrelevant to their dissertation, then go off on a sidetrack and do lots of reading that is not helpful.

Perhaps less often:

1. They are slow to narrow down their topic so they waste time reading widely when they should have better focus.
2. They don't know how to choose the most relevant literature.
3. They read poor quality material (e.g. blogs).

A friend doing a doctorate complained about lack of journal articles on her topic. However, within (literally) a minute, I’d got a list of about 1,500 refereed journal articles. Even if only 10% are useful, that is still a lot of reading.

Many items are free on the internet and there are many websites. These are my favorites:

1. Google Scholar is an excellent way to search for articles. (It is not the same as regular Google.)
2. I'd add Core to my essentials list. http://core.ac.uk
3. In education, it's hard to go past ERIC (Education Resources Information Center)
4. Try Directory of Open Access Journals

Other than that, if you know of a title that looks promising, use Google advanced search. There will probably be a copy for free somewhere.

Learn how to use key words to focus your search. In fact, most journals require authors to list key words so that articles will be easier to find.

If you need to search for a specific phrase, try Google advanced search, which is not the same as the regular Google or Google scholar. This is especially useful if you find something in the references of an article you found. Many articles are behind a paywall, but are often also somewhere else on the net for free. (For some reason, these publishers try to get payment for articles that are no longer recent.)

Search engines tend to list search results in this order:

1. If they are commercial, the top of the list is paid ads that are relevant to your search, and hits for items that have paid to be near the top of the list.
2. Hits for items that are a very good match for your search criteria.
3. Other hits that are not such a close match, I.e. fewer of your search keys.

First, how comprehensive and up to date is the literature review? Expectations vary between university departments.

Second, how well-supported is your choice of topic? Does the current literature demonstrate a need or a gap?

Third, do you have enough to support your conclusions? Is the evidence firm? Would an alternative or contrasting conclusion still be plausible?

If your review covers a wide enough breadth of the literature, you will be able to identify biases and assumptions quite easily.

You can and normally should include all serious research on your topic. You seldom need to avoid biased research unless it is quite wacky. You should give a critique that identifies any biases. This is especially important in history, where everybody comes with particular perspectives.

Yes, you can. Abstracts usually come with the full publication details (for your bibliography), giving you three options:

1. You have enough enough information for a brief mention of what the writer did and the results.
2. It might be adequate to refer to the paper as a Cf. or See also.
3. You’ll need the full text if you need to do much analysis or critique. You could search the internet for a free copy because it's probably somewhere for free. Otherwise you'll have to pay for a copy.

First. it's true that the volume of literature is now overwhelming.

There are now several systems to manage the mess:

1. The main kind of system is the major free agglomeration services, and the best use of them might be all you need. Google scholar is the most well known. Pubmed for medicine, Eric for education. Core is bigger, and there are now dozens of others. Do a search using different combinations of key words and you'll find lots of stuff. To make this easier, journal publishers now normally require writers to include keywords in articles.
2. A couple of other agglomeration services are available by subscription. Ebsco is one of the best known. These are valuable for resources that are not on the internet anywhere else.
3. To keep track of everything, a couple of numbering systems are in use. ISSN and ISBN have been around a long time. More recently, doi numbers identify individual items. There are also numbering systems to identify individual writers; orcid is the most widely used.
4. There are now pre-prints. These are unpublished early versions of works that will be published.
5. Open access means that readers do not have to pay for items. This makes them much easier to get, and good open access journals are listed in the agglomeration services. (Some categories of open access also allow others to re-post items elsewhere.)
6. Many institutions now place theses and dissertations in repositories, some of which are agglomerated. Even if not, they can often be found on general internet search engines. Use Google advanced search with .edu in the domain name slot.
7. While you can't use Wilipedia, you can chase up the footnotes of relevant articles.
8. Anytime you find something good, you can find more items in the references.
9. Many academic libraries are now connected through Worldcat.

💡 It is usually easier to write an annotated bibliography first. It is then easy to edit into a literature review.

Warnings

• ⚠ Online agglomerators probably won't help you find many books and journals published only on paper.
• ⚠ If you find a reference to an article that seems relevant but the only access is to buy a copy, many prices are exorbitant. Try a google search; there's probably a copy out there for free somewhere.
• ⚠ If you find an article and lose it somehow, it can be nearly impossible to find again. Download useful items when you locate them.
• ⚠ While a dissertation literature review must be comprehensive, this is defined differently in different institutions. I also wonder if people ever find everything relevant; how do you know that you've missed something relevant in an obscure journal article or poorly circulated or out of print book?

The relevant criterion is not length but whether it gives an accurate and comprehensive picture of the current state of research on the topic. Consequently, there is no minimum or maximum length.

In most cases this requires lots of work and a PhD dissertation writer should expect to write anything from 30 to 60 pages. (In some cases it can even be much more.) A few topics require very little or nothing, but they are quite rare.