The writing is an essential skill to share with the scientific community the findings obtained by the researchers; however, it requires the implementation of a series of strategies whose domain has not yet been acquired by researchers with less experience. This research has a double objective; on the one hand, it tries to know the opinion of the PhD students about their own process of scientific writing and, on the other, it analyzes the scientific texts produced by the novice researchers. For this, two instruments have been created: a brief questionnaire and a system of categories to analyze the 397 errors that have been detected in the scientific texts written by 11 PhD students from different areas chosen for convenience. The results show that approximately 40% of researchers with little experience previously plan the texts they write and, for the most part, use a scheme for this. In addition, more than 90% believe that the most complicated thing is to provide coherence and cohesion to the scientific text. In relation to errors, the most common occur in the introduction (43.6%) and the method (18.4%) of the text and are grammatical (39%) and APA regulations (24.9 %). Finally, some limitations of the study are discussed.