These are questions dealing with attitude and are the most important questions when doing qualitative social science research to gauge relationships among events. In addition to construction questions about attitudes, it is important to have the questions drafted in the correct format (Nachmias, 2008).
The Quantitative methodologies will be the statistical tests designed for the overall model to incorporate the information provided through one, two or all of the Qualitative data analysis methodologies. The tests used to determine the relationship between these "qualitative" factors and increases in Infection rates, will be the Chi-Square, Student's T-Test, ANOVA (to test for variations among the data), the construction of a Linear Regression Model and the calculation of the Pearson Correlation Coefficient, otherwise known as "R-Squared" (Nachmias, 2008).
These tests will be utilized in conjunction with a predetermined level of significance, or alpha. Since these tests will all be measuring the means and relationships of one data set, the level of alpha to be used in these tests is set at .05. Therefore, the resultant calculations will be compared to this level of alpha to determine if there is any statistical significance between these factors and increases in infection rates. To ensure the validity of the linear models, the R-squared value will be calculated and included within the analysis. There are numerous classes of the R-squared value, however for our purposes; it will be the "coefficient of determination" that accompanies linear regression models. Consequently, the main role of R-squared in our analysis will be to provide an explanation for the variances in the data. For example, if the R-squared value of the PDI for a certain subset of the population equates to .88, this would translate into the PDI being responsible for 88% of the data results, with 12% being attributed to chance or randomness. The formula that will be incorporated into this research for R-squared is as follows:
and the traditional standard of "goodness of fit" will be applied to the R-squared values; wherein, a R-squared result of 1.0 indicates the data fits perfectly and conversely, a R-squared value of 0.0 indicates no correlation between the two variables and all the data points are the result of randomness and chance. Furthermore, the various Chi-Square analysis will be broken down across age categories in order to determine the impact of generational status on the various elements. The standard formula for Chi-Squared will be utilized; the formula is as follows:
The allowable error margin as plus or minus 15 percentage points.
Pattern of responses
The expected pattern of responses is as follows (see plot). 'Response a' (40%), 'Response B' (60%). In particular, the percentage for 'Response a' is 40%.
Margin of error
One factor that determines the required sample size is the acceptable margin of error. If we are willing to accept a relatively wide margin of error we'll need a relatively small sample. By contrast, if we desire a relatively narrow margin of error we'll need a relatively large sample.
For an error margin of plus/minus 15.00 points we need 41 subjects. If we were to double the error margin (to 30. The sample size would be reduced to 11 subjects. By contrast, if we were to cut the error margin in half (to 7. The required sample size would increase to 164 subjects.
We assumed that the percentage in this category is 40% which led to a sample size of 41. If the true percentage is actually 30% the required sample size would be 36. If the true percentage is actually 50% the required sample size would be 43.
In computing the sample size we assume that we want to be 95% certain that the observed value falls within the margin of error (rather than 90% certain, for example) and also that we are concerned with errors in either direction. Changing either of these assumptions would also affect the sample size required.
According to the data presented in the Figure, the R-squared value is .7987. As this graph demonstrates there is a distinct correlation between these two variables. The dotted lines represent the Confidence Interval for this analysis. This Confidence Interval represents the relationship between these two variables. The solid line in the figure represents the regression of the means of the data involved in the analysis. As the Confidence Interval is closer to the solid line this indicates that the relationship is based on a statistical relationship and not random chance. As this figure demonstrates, the relationship between washing hands by nursing staffs and the increase in Nosocomial Infection rates is based on statistical correlation, therefore it can be asserted that this one specific, cultural value can have an impact on attitudes and behaviors and negatively impact these behaviors to the end result of increasing infection rates.
Using the standard Chi-Square Test formula we arrive at the following. The p-value for this Chi-Square Test is .0001 or there is a .01% chance that the variations within this data are due to chance and randomness. Therefore, it can be concluded that Hand Washing Protocols are a very significant factor in developing various levels of Nosocomial Infections. Utilizing the Standard T-Test equation, we arrive at the following. The t-value is .3171 with a corresponding p-value of .7619. Since the p-value is greater than the predetermined level of significance, .05; it can be concluded there is a 76% chance the variations in this data are predicated on statistical relationships.
With the p-value of the T-Test being in doubt it is relevant that another form of analysis be conducted to determine if there is a relationship. Therefore, a linear regression model with R-Squared testing was conducted. Using the same data we arrive at the following result. The R-Squared value is .9026. This indicates the relationship between Hand-hygiene and Infection Rates has a 90.26% correlation. Therefore, this indicates the strong relationship between the two values. An additional test conducted to ensure validity was an F-Test. Using the data from Table 2.8, the F-Test revealed F = 151737, the P. value is < 0.0001. This test suggests that the difference between the two Standard Deviations is extremely significant.
95% confidence interval
Error from to
0.007777 -0.007922 0.05901
0.2925 1.168 3.686
As the table demonstrates, the 95% Confidence Interval is between -0.007922 and 0.05901. The slope o the linear regression is .02554, this is within the Confidence Interval of the model. Furthermore, the R-Squared calculation is 0.84; therefore the data represents an 84% relationship between Hand-Hygiene and Infection Rates. As we can see, the Linear Regression (represented by the solid line) is relatively close to the upper and lower limits of the Confidence Interval (represented by dotted lines). This indicates there is a likely correlation between the Mean MAS scores and reliance on We project the likely findings of such a study based on the findings produced by similar studies of this nature. Accordingly, the study by Sax et al. is compelling in terms of the preeminent finding the social pressures are likely to have a significant impact on handwashing behaviors. The consideration that behavioral tendencies in this area will hinge upon normative conditions proceeding from peers, superiors and the industry as a whole does provide the research with an area from which to draw some clear recommendations. Namely, as the discussion here below touches upon the failure of certain training objectives with respect to hand hygiene, it is useful to consider that an entwining of hand washing goals with other cultural priorities can significantly impact the positive adoption of desired behaviors.
This finding would be reinforced by an echoed perspective in the study by Barrett & Randle, which similarly identified the capacity for social pressure and the construction of a positive culture endorsement of hand hygiene as having a likely positive impact on behavioral tendencies in this area. To this end, they would observe that "respondents emphasized the importance of fitting into the clinical area and role models in shaping hand hygiene compliance." (Barrett & Randle, 1851)
Infections related to healthcare are among the most important causes of morbidity and mortality in hospitalized patients. A study of prevalence carried out by the World Health Organization (WHO) in 55 hospitals from 14 countries, showed that 8.7% of hospitalized patients contract Nosocomial Infections (NI). The importance of NI in terms of morbidity, mortality, impact on quality of life in patients and relatives and secondary economic costs, has been emphasized repeatedly in the last years . In the developed countries, around 5-10% of patients admitted to hospitals for acute conditions presented an infection that was not being incubated or present at the time of admission. Healthcare-related infections are the direct cause of 80,000 deaths in the United States and 5,000 deaths in England every year. According to data from the Survey on Prevalence of Nosocomial Infection in Spain (EPINE study) for 2006, NI affected between 7% and 9% of patients…