On the basis of the reference period, the study designs can be categorized into retrospective, prospective and retrospective-prospective study designs
The aim behind retrospective study design is to study some event, phenomenon or situation that has been happened previously. There are only two ways to collect data for a retrospective study: either the investigator collects information from written evidence like from books, magazines, newspapers, diaries and other personal records, or he asks the respondents who can recall the situation. Asking the respondents about the situation is only possible if the event has happened in a certain time period in the past if the event happened several decades or years back the investigator cannot get the data from people rather he has to rely on written records.
The prospective study design studies the impact of a phenomenon, situation, program, attitude or problem in the future. Most of the studies that are conducted using the prospective study design are also classified as experimental studies since the aim of the investigator is to wait for the impact of a program or a situation to see the results on the target population. Unlike retrospective studies, the prospective studies use experiments in the field or laboratory to explore the outcomes later on. The outcomes that are established as a result of these study findings can then be generalized for the future happenings of the same kind.
Retrospective-prospective studies focus on a phenomenon in the past and study it in the future. The only difference between these studies and before-and-after studies is that they do not have a control group. Most of the before-and-after studies without a control group are retrospective-prospective study designs. In retrospective-prospective study design, the investigator collects some data about the population before the intervention or treatment is applied and then he studies the same population after the treatment of the intervention has done its impact on that population. The investigator does not analyze the population at the baseline neither he has a control group.