0ver recent years research has been an integral part of nursing, and has helped much in the development of health care delivery and interventions, with concurrent changes in nurse education. Nurses nowadays are more research-oriented and are expected to integrate the findings of various relevant studies in their work to better the quality of care provided for patients. However, Polit & Beck (2004) emphasise that research findings should only be used in practice following critical review of the relevant studies to ensure the evidence is sound and appropriate. Thus, research literature needs to be regularly reviewed and evaluated through critiques so as to gather evidence on a particular topic and to develop a balanced evaluation of a study's contribution to knowledge. Burns & Grove (2005) would add that the critique of studies is a vital process in the development of knowledge for use in practice or as a baseline for other studies and lay emphasis on the meaning and purpose of critiquing a research, and will therefore define it as a systematic, unbiased, careful examination of all aspects of a study to analyse the credibility and limitations based on previous research experience and knowledge on the topic and not just barely criticizing in the negative aspect which is very often linked with the word 'critique'. Miller et al (2006) study on 'substance use, psychiatric symptoms and the onset of schizophrenic illnesses has been selected for 'critique' in this essay. The aim Miller et al's study is to investigate the effect of continuous use of cannabis in the lifetime of chosen individuals upon development of psychosis and schizophrenia. The researchers used a quantitative research method. This method does not describe subjective experiences, but instead involves analysis of numerical data (Parahoo 2006). It has higher controls over the research variables compared to qualitative methods, and its findings can usually be generalised, depending on the sampling method used (Nieswiadomy 2008). Positivist theorists claim that quantitative research methods are important because they provide objective basis for evaluating claims of causal relationships. Based on the probability theory, it provides a superior measure for evaluating claimed causal relations, and for choosing between two or more competing theories (Creswell 2008). However, it must be noted that the choice of a quantitative or qualitative approach depends on the research question being investigated. Miller et al (2006) have appropriately selected a quantitative study on the basis of their research questions. Rees (2003) underlines the importance of critiquing a piece of research in asserting that its purpose is to make sure that the service user gains the maximum knowledge available in a particular nursing field and also lay emphasis on the fact that in order to successfully evaluate an article a systematic framework for critiquing should be used. To assess the relevance and accuracy of the research, the critiquing tool adapted from Carlson et al (1999) was used as it provides a comprehensive checklist of specific questions to assess the research method. These questions were used to evaluate the quality of the study and thus give an indication of the reliability of the findings.
Research objectives, questions and hypothesis are formulated to bridge the gap between a stated research problem and purpose and the detailed design and plan for data collection and analysis but the only difference is that hypothesis can predict the outcome of the study (Burns & Grove 2005). However the researchers did not formulate any hypothesis for their experiment, only the research purpose and questions are present and these were generated from the problem identified, i.e. the high intake of cannabis by individuals from 11-25 years old at various times during their life upon the development of the psychotic symptoms of schizophrenia. Hypothesis are generated by observing phenomena or problems in the real world of nursing practice, analysing theory and reviewing literature (Burns & Grove 2005).The appropriate hypothesis for this study would have been for example "the more cannabis an individual takes in his lifetime the more likely he is to develop schizophrenia".
Measurements and time frame
The subjects for the study were assessed for clinical features, neuropsychology and brain structure through structural magnetic Resonance Imaging (MRI). The mental state of the subjects was also assessed at entry and regular follow ups were made nearly every 18 months using the Present State Examination (PSE). The PSE itself was divided was divided into 12 subscales of symptom detailed by Owens et al (2005). The time frame for the study is not stated in the article and the reader will have no idea when the research did exactly start. However, it is mentioned that some results were produced over a period of ten years of the study which can imply that at least ten years have been done on this particular subject. Moreover, it is not mentioned that the research ended or continued after ten years. However, MRI and PSE are both tools that have limitations e.g. MRI is time consuming and may not always be able to tell the difference between some disease processes (VMC 2009). The expertise in examining the results obtained from the MRI is not stated anywhere in the study and misuse of it or misinterpretation of the MRI scans could cause bias to the outcome of the study.
Sample and Setting
The number of subjects selected for the study and the sampling method used are not clearly stated, making it difficult to follow the study. The researchers stated that three groups are used for the study: 1 experimental group and 2 control groups. They then state that they planned that the two control groups would consist of approximately 35 subjects which according to the researchers was the required amount of high risk sample predicted to develop schizophrenia, but the number of subjects eventually selected for the study is not only confusing but the rationale for using what seems to be a much higher number are not given. The same argument applies to the characteristics of the high risk group as the only thing mentioned about the subjects was their age, i.e. between 16 25years old. The sample were chosen throughout Scotland, mainly from places where extended families still existed as according to the researchers, they provide the ideal network for the study. But this explanation does not make the readers any wiser. The subjects for the first control group were chosen from the social network of the subjects of the high risk group whereas the subjects for the second control group were chosen from local hospitals. From the article it is mentioned that the subjects were chosen mainly throughout Scotland, and presumably this is the context where the study took place. It is essential in any study to get an appropriate sample size that will generate enough data (Auerbach & Silverstein 2003), and the sample population must reflect the characteristics of the target population so that the findings can be generalised (Parahoo 2006). Sample size considerations are essential in any study and if the methods used are not appropriate, then researchers are more likely to misinterprete findings due to inappropriate, unrepresentative and biased samples (Devane et al 2004).
The study does not give any information about the qualification of the authors and it is therefore not possible to assess how closely they are involved to the subject matter or to establish their accreditation as researchers though it is mentioned that the study was carried with the approval of the Ethical Committees of the Relevant areas and family members of the experimental subjects. The expertise and knowledge of the authors can add value to a study (Cluett & Bluff 2006) but it must also be recognised that all research is potentially exploitive and researchers' motives can often be mixed (Shamoo & Resnik 2009). It may be presumed that the2 authors were health professionals, a point, which can lead to a 'blurring' of roles, goals and motives and it is important to ensure that ethical implications are considered while the study is being done. Parahoo (2006) would mention that researchers should ensure that the need to obtain data should not precede patients' needs, wishes and rights. Abbott et al (2001) would also mention that it is of prime importance that the subjects and researchers have their human rights protected but nothing about the consent of the subjects to the experiment and their freedom of choice as a subject is mentioned in the article.
Lodico et al (2006) would mention that in any experimental study there are many possible extraneous variables also known as cofounding variables (Parahoo 2006). These are variables that are mistakenly manipulated along with the independent variable. Researchers should be aware of these variables and should aim to control them so that they do not get involved with the experiment consequently damaging the internal validity of an experiment (Parahoo 2006).In this article, two such variables were identified namely: Childhood Behavioural traits and intake of illicit drugs other that cannabis e.g. alcohol intake and cigarette smoking. For the subjects aged between 13-16 years old, their mothers were made to fill in a Childhood behaviour Checklist (CBCL), (Achenbach 1991) which contained 120 items each scoring a three-point scale and from the list of items , nine summary scales were derived to well define the usage of cannabis. Another scale was used for the use of illicit drugs where alcohol intake and cigarette smoking were given recommended weekly limits for the study. Two control groups were used together with the experimental group. The subjects for the first control group were chosen from the local hospital and subjects for the second control group were chosen from the social network of the high risk group.
Intervention and Data Collection
Subjects were assessed in terms of clinical features; neuropsychology and brain structure was determined by structural magnetic resonance imaging (MRI). The mental state of the subjects was assessed by making use of the tool, Present State Examination (PSE), tool devised by Wing et al (1974). The baseline clinical features included childhood behavioural characteristics and it was studied using the Childhood Behaviour Checklist (CBCL), Achenbach (1991) which the mothers of subjects aged between 13-16 years old had to fill. There are 120 items on the list and each one of them scored a three-point scale. From these items nine summary scales were derived and the researchers also made use of another scale for the use of illicit drugs other than cannabis. The CBCL is quite a complex tool for people who do not know how to use it properly. It is stated that the mothers of some subjects had to fill the form for their child, but it is not mentioned that these form were properly filled by the mothers or if were checked for errors by the team of researchers. It is an important factor to consider as Child behaviour was a cofounding variable which if it was not controlled and checked could have cause bias to the outcome of the study. Moreover, the results obtained from this study do not tell us how that study was done meaning of how the observations were done. Researchers should provide adequate information on how observations were carried throughout the study (Parahoo 2006). For example, were the subjects aware that they constantly being observed? If so, this may have had an effect on the outcome known as the 'Hawthorne effect.' These are due to changes in subject behaviours, on being involved in a trial, due to increased knowledge or interest, or due to feeling observed (Braunholz et al 2001) and this would eventually cause bias affecting the outcome of the study. Ideally, to reduce the occurrence of bias in the study, the experimental subjects and the person assessing the outcome of the study should not be aware that they are being assessed.
Data description and analysis
Logistic regressions and t-tests were used as statistical test to determine the different values and were well represented in tabular form. However it is mentioned in the article that the twenty one subjects that developed schizophrenia were not followed up after illness onset. Were these subjects removed from the experiment and if so how was the sample size balanced again so that the outcome of the experiment could not be affected. It is also mentioned that the data of the subjects with three or more testing occasions were restricted to two or three testing occasions but it is not clear how this could help in the experiment. Burns & Groove (2005) would mention that the research objectives questions and hypotheses will help to determine the choice of analysis techniques to be implemented, but since this study lacks hypotheses, we do not know how wrong or right the researchers clear about the test they were going to use to generate the results for the experiment.
All the results obtained after the different tests were done were put in tabular form and their meanings were then explained. It is shown that the inter-rater reliabilities (Pearson r) value for cannabis was 0.95 and that of cigarette was 0.91 compared to alcohol which was 0.88. Cannabis and cigarette values are relatively high but in the article it is mentioned that other substance use variables did not show any other significant relationship with the illness onset. If this was true why would cigarette smoking yield a value close to the one from cannabis use and how the better results were produced without taking this value into consideration. It is the same for category of other drugs which the researchers chose not to take into account. The researchers claim that their study did show an association between frequent use of cannabis and the onset of schizophrenia. They also reported that their hypotheses were supported, which as stated earlier were not mentioned of in their research methodology. On the whole, their claims are generally not supported by the statistical findings as well as their subsequent discussion, which on the whole show many statistically non-significant results. Indeed, the authors appropriately compare the findings of their studies to various comparable studies, and admit that the associations they found between cannabis use and schizophrenia are somewhat more tenuous that those reported in the other studies. The results of the study must also be judged against the various limitations (e.g. unreliability of self-reports) which the researchers themselves have highlighted.
The research used as evidence
When the results from this study were compared to other studies, it was noted that more attempts should have been made to investigate the link between the frequent use of cannabis and the development of psychotic illnesses in the high risk subjects with a familial vulnerability to psychosis. The present study showed that only one attempt has been made in this filed (Verdoux et al 2003). However, it is stated in the article that the objectives of the study revealed that cannabis used at various stages in the lifetime of the subjects predict the development of schizophrenia within high risk group. Arseneault et al (2004) would agree to this in saying that cannabis use has an impact on the development of Schizophrenia but his argument was in contradiction with another study made by Verdoux et al (2002). However, there was a difference in the sample used for both studies and it is mentioned that there were no cannabis use monitoring after intake and also that the cannabis use from the subject of the other study was low compared to this one. The relationship within the high group risk sample was also noted to be different in the samples for this study. Behaviour problems once eliminated in some subjects showed that these subjects took less cannabis later on during their life and hence reduced the outcome of schizophrenic condition. It was also mentioned that heavy use of other illicit drugs such alcohol and cigarette smoking had no significant effects once childhood behaviour and cannabis use was control but it was not mentioned if these illicit drugs could have an effect on the onset of schizophrenia if the childhood behaviours were not controlled or taken into account in this study. This study shows through the different tests done that there is possible link between cannabis and the schizophrenia but it has not been put forward straight away that cannabis is directly or indirectly the cause schizophrenia in high risk people or normal people. This study would have more convincing if the high use of cannabis taken by the subjects at risk was only taken into account and not be associated with other external factors like childhood behaviours. The interaction between use of cannabis and medication should have been taken into account as well because medication sometimes induce psychosis and how to distinguish between psychosis cause by medication or cannabis would have been an interesting point to consider in the study. There are some areas in this article which require further development but nevertheless it shows evidence about the likelihood of an impact from frequent use of cannabis in the life of an individual upon development of schizophrenia and can thus our practice can be base on the findings of this research study.