You should begin with a descriptive analysis for the survey rating question “All of the practice changes so far have been practical and fit well with the workflow of the unit.” This can be done with the Distribution platform. The JMP output is shown in Figure 2.5 JMP Distribution Output for Response to the Survey Question “All of the practice changes so far have been practical and fit well with the workflow of the unit”.

To round the proportion in each category, right click any of the values in the Prob column, select Format Column and choose the appropriate number of decimal places.

Notice that 47 (18%) of the nurses chose not to give a rating. The proportions given in the Frequency table are based on the number of nurses who responded to this particular question.

Neutral is the most frequently occurring response with 42% of the nurses selecting this rating. The histogram and the frequency table present the same information. When reporting these results, choose the presentation format that will be most easily understood by your audience.

A descriptive analysis of the respondents’ demographics will provide additional insight into the survey results. A table of descriptive statistics can be created with the JMP Tabulate platform (Analyze > Tabulate). Highlight the four columns containing the demographic variables (Primary Role, Years of Service, Employment Status, and Education) and drag them to the drop zone for rows, then highlight N and % of Total from the column of statistics and drag them to the drop zone for columns. Adjust the format so the % of Total is rounded as shown in Figure 2.6 Tabulate Dialog to Create Table of Nurse Demographics.

Figure 2.7 JMP Tabulate Output for Nurse Demographics shows the resulting table containing the frequency distributions for each of the demographic variables.

We see that 88% of the nurses are registered nurses, there is a fairly uniform distribution of years of service with the hospital, 83% of the nurses are employed full time, and 60% hold an associate’s degree. A descriptive analysis is a necessary first step in your data analysis. It is important to give your audience a statistical summary of the data you have collected.

The proportion of nurses that view the current state of EBP favorably would correspond to responses for the agree and strongly agree ratings. The neutral, disagree, and strongly disagree ratings can be combined to indicate those nurses that do not rate EBP favorably. Create a new column, “EBP Favorability” that combines the agree and strongly agree responses into a “Favorable” level and the remaining responses combine into the “Not Favorable” level. This new column can be created using the JMP recoding feature. Before using the Recode command, you will need to remove the List Check option from the column containing the responses to this survey question. For this recoding, we want to create a new column called “EBP Favorability” by highlighting the column “All of the practice changes so far have been practical and fit well with the workflow of the unit” and selecting Cols > Recode and choosing “New Column” option, as shown in Figure 2.8 Completed Recode Dialog to Create EBP Favorability Column.

We do not want to overwrite the original 5-point Likert responses as they may be needed in a future analysis.

The JMP Distribution platform allows us to summarize the EBP Favorability column and compute a confidence interval for the proportion of nurses who view EBP favorably. The frequency distribution for the EBP Favorability column is shown in tabular and graphical form in Figure 2.9 JMP Distribution Output for EBP Favorability. The mosaic plot was obtained from the red triangle menu.

The histogram shows more detailed information than does the mosaic plot. The mosaic plot may be a preferable visualization when there are space limitations or a large number of nominal or ordinal variables to summarize. Of the nurses that responded to the survey, 33% have a favorable opinion of the current implementation of EBP changes in their unit.

A confidence interval for the proportion of nurses who view EBP favorably can be obtained from the Distribution platform by selecting Confidence Interval from the red triangle menu associated with the EBP Favorability variable. A 95% confidence level is chosen for this case. Figure 2.10 JMP Confidence Interval Output for EBP Favorability shows the JMP confidence interval output.

The best estimate of the proportion of nurses that view EBP favorably is 0.33. A 95% confidence interval for the proportion of respondents who view the EBP implementation favorably is [0.27, 0.40]. There is 95% confidence that the true proportion of respondents who view EBP changes at the hospital favorably lies between 0.27 and 0.40. The 95% confidence level means that on repeated sampling we would expect 95% of such intervals to contain the true proportion. Different confidence levels can be chosen, but 95% is commonly used in health care applications. The confidence interval takes into account the precision with which the sample proportion was estimated.