The UK Ministry of Justice has completed a literature review about self-represented litigants (whom they call “Litigants in person.”) The review included international sources, at least as to formal research.

The overall conclusions:

    While some good quality evidence existed, this was limited. Together the studies in this review provide useful indicators of the types of motivations, problems encountered and outcomes for litigants without legal representation (litigants in person). However, there are still a number of gaps in our understanding of this issue.
    The term litigant in person covers a range of scenarios. Individuals may have received varying degrees of legal advice; may have chosen to litigate or had claims brought against them; and may or may not have themselves participated in proceedings. One UK study suggested unrepresented litigants in family and civil cases were common.  Most unrepresented litigants were inactive, particularly in civil cases.
    It appeared litigants in person tend to be younger, and have lower income and educational levels, than those who obtain representation. Suggested reasons for lack of representation included funding difficulties and the belief that cases were simple enough to be heard without a lawyer.
    Litigants in person could face problems in court, such as understanding evidential requirements, identifying legally relevant facts and dealing with forms. It was suggested that the oral and procedural demands of the courtroom could be overwhelming.
    Research with other court participants, such as court staff, the judiciary and other parties’ representatives, suggested they felt compensating for these difficulties created extra work and possibly presented ethical challenges.
    The evidence on the impact of litigants in person on case duration was mixed. This appeared to be influenced by how active the litigant in person was and by case type. The evidence suggested cases took longer when the unrepresented litigant was active and could take less time when the litigant was inactive. Some studies found that family cases without representatives were less likely to settle, increasing case duration.
    The weight of the evidence indicated that lack of representation negatively affected case outcomes, although few of the studies reviewed controlled fully for case complexity. This was across a wide range of case types. There were indications that in some cases specialist lay representatives were as effective as legally qualified representatives.
    A number of studies investigated assistance for litigants in person, presenting positive findings on litigant and court staff satisfaction where such assistance was received. There was little research examining the impact of the various methods of assistance on case outcomes.

This is the section on innovations:

What action works in assisting litigants in person?
There were a number of examples in the literature of ways in which litigants in person were assisted. These ranged from studies on court-based advice services, self-help and hotlines (Morris et al., 2006; Plotnikoff and Woolfson, 1998; Henschen, 2002; Goldschmidt et al., 1998; Philliber Research Associates, 2009; Judicial Council of California, 2003 and 2005; Pearson and Davis, 2002; Greacen Associates, 2008). Most of this is evidence from the US.
As both Engler (2010) and Greacen (2003) noted in their reviews, although the evidence on the whole indicated high levels of user satisfaction with services and that court staff appreciate them, there was little evidence of the impact of these efforts on court outcomes. Adler (2008a and b) examined the effectiveness of pre-hearing advice for tribunal cases. He found that, for some types of cases, people who received pre-hearing advice but represented themselves did almost as well as those who were represented. However this was the only study identified that examined this issue.
Additionally, some studies identified ways in which judges assisted litigants in person, and adapted their style to help them through the court process (Goldschmidt et al., 1998; Greacen Associates, 2008; Genn and Genn, 1989; Moorhead and Sefton, 2005; Kelly and Cameron, 2003; Hunter, 1998; Lewis, 2007). Judges sometimes adopted a more interventionist role to compensate for the difficulties of litigants in person, questioning them to elicit evidence. They also sometimes altered the order of proceedings to help those without representation, and provided explanations to litigants. There was, however, little evidence of the effectiveness of these efforts. Some studies indicated that there was some worry among judges about the propriety of adopting a more interventionist style and that they might appear to be unfair (Genn and Genn, 1989; Goldschmidt et al., 1998).

It should not be forgotten that this review was done against the backdrop of very severe cutbacks in legal aid in the UK.