Statistical Analysis and Nurse

¶ … health care centers (PHCC) in Stockholm County, 40 of them were randomly selected using an old-fashioned, non-probability method of basically drawing names from a hat. The author notes, "every PHCC was given a unique number that was written on a paper card and placed in a pot. For transparency, two colleagues independently drew 20 paper cards each, a total of 40." Of these 40, one declined to participate. Therefore, 39 PHCCs were selected, and one nurse from each PHCC served as contact person. The sample size is adequate and actually fairly large for the study. Although unconventional, bias was not introduced by using this method of sample selection, and the sample can be considered representative of the population given the randomness of the PHCC selection procedure. Eligibility criteria are also clearly identified, as the contact person nurse needed to comply with the study design, namely to distribute anonymous questionnaires to colleagues. Informed consent was acquired during the selection procedure, and the rights of the subjects were insured by coding the questionnaires.

7. Data Collection

Data collection instructions are clearly identified, and include the initial questionnaire and follow-up interview. These methods are appropriate for the problem being studied, which was to measure preparedness. All subjects received the same questionnaire, so there was no internal validity problem. The instruments were designed specifically for this study, to measure exposure to intimate partner violence. Instrument validity is limited, but the questions are clearly outlined to ensure test reliability.

8. Data Analysis

The researchers use STATA 9.0 for statistical analysis and descriptive statistic acquisition. Frequency tables were garnered from the software. A Pearson's chi-square test was used with p-value of

9. Analysis of Findings/Discussion

As exploratory research, there was no directional hypothesis. The researchers set out to discover and assess nurse current level of preparedness to identify intimate partner violence issues and work with women exposed to intimate partner violence. With the results of the research, healthcare administrators may choose to make changes to their policies, or nurse educators would be able to include intimate partner violence awareness and training into curricula.

The results of the current research are interpreted at face value, as no directionality or causation can be offered. The context and severity of the problem is discussed, as well as the theoretical framework within the nursing field. Specifically, the researchers identify organizational and structural impediments to helping patients with intimate partner violence. Training issues are also identified, as most of the nurse participants did not know how to ask their patients about intimate partner violence, let alone advocate on their behalf. Therefore, generalizations are made about the current state of preparedness and what can be done at the administrative level to remedy and improve the situation. There are no unwarranted interpretations of causality, as the authors do not presume any causal relationship such as between nurse preparedness and patient outcomes. There are a few graphics to accompany the text and highlight the key findings and major variables. These graphical elements improve the overall usability and reader experience.

10. Conclusions, Implications, Recommendations

The results of the study are summarized concisely, even though the researchers include a wide range of variables in the questionnaire. Comparing their results with prior literature, the authors found that few nurses are adequately prepared to question patients about intimate partner violence, or inadequately trained to recognize some signs of intimate partner violence. About half of participants stated that they directly ask about intimate partner violence if it was suspected, which the authors claim to be higher than in previous research.

The authors openly discuss the study limitations, including wide confidence intervals in the regression models and potentially low statistical power of the results. Yet response rates were generally high and the sample size was large. Because some nurses declined to participate, the authors note that there may be a selection bias due to the sensitivity of intimate partner violence as a subject matter. Nurses who declined to participate might, for example, have been concerned about patient confidentiality or their own confidentiality.