Potential problems that could arise during the funding stage are: the evaluation objective process is designed to ensure that eligibility criteria are applied consistently and impartially to reduce the risk of errors or distortions in the synthesis of evidence. Articles identified during the research are generally structured to include a title, summary (or summary) and/or a «full text» version such as a scientific newspaper, agency report or website. The eligibility test can be applied to these different reading levels in order to impose a number of filters with increasing rigour, and screening is therefore generally a gradual process. CEE recommends using at least two filters: (i) a first reading of titles and abstracts to effectively remove articles that are clearly irrelevant; and (ii) evaluation of the full version of the article. It may be difficult to determine whether articles are related to each other, since related articles do not always cite each other (Bailey 2002; Barden et al., 2003) or common authors (Gétzsche 1989). Some «detective» work (for example.B. Check whether the same data appears in more than one article or contact authors) may therefore be necessary by the audit team. While it is ideal to identify related articles that relate to the same study at an early stage, it is only during the full screening phase that one can realize that the articles are related. Once the links between the articles and the studies have been established, all articles related to each study must be clearly recorded. This can be done using a separate document or table, or using grouping or sponsoring functions that can be used in reference management library tools. The screening procedure is carried out jointly by the European Commission and each of the candidate countries so that they can become familiar with EU law and demonstrate their ability to implement EU law.
In order to assess the consent of the screen, it is necessary to verify a sample (as large as possible) of the articles identified during the search by at least two people and to determine their consent. Sample size should be justified by the audit team and articles that include the subset should be randomly selected to avoid distortions with respect to certain authors, themes, years or other factors. An evaluation of the agreement reached between the farmers during the pilot tests may contribute to the reproducibility procedure being reproducible and reliable (Frampton et al. 2017). If necessary, the eligibility criteria and/or screening procedure may be modified and retested to improve the agreement between the screeners, provided that the derogation from the protocol is explained and justified. The agreement can be assessed by: recording the observed proportions of items for which the pairs of screens agree or do not coincide on their promotional decisions; Calculating a statistic of revision agreements; and/or how to describe, not express themselves and discuss differences of opinion. A dispute resolution procedure between screeners should be agreed by the audit team to ensure consistency, which should be established in advance in the protocol.