Research Questions, Hypotheses, and Statistical Tests: The ABC Hospital
The purpose of this assignment is to define research questions in relation to their corresponding hypotheses, using a dataset of various variables. The statistical tests that follow ensure that degrees of significance are necessary to assess whether the satisfaction survey can help decrease patients complaints at ABC Hospital. The writer/researcher uses the completed table1 from SLP1 to analyze the variables in relation to table 3, to better appreciate the statistical assumptions that the magnitude of the differences in the variables considered are minimized. Various statistical tests such as t-test, one way ANOVA, two way ANOVA, and Chi-square will be performed. For one way ANOVA, it is plausible that the writer/researcher will use only one independent variable in the analysis. Such variable may have many levels to address the questions under study, and tests the null hypothesis (Meyers, Gamst, & Guarino, 2006). Several ANOVA summary tables will also be presented to assess the independent variables, based on the data provided. A one way between subjects ANOVA to assess the between-subjects variance is essential.
The writer/researcher thinks it would be asinine not to look at some univariate outliers pertinent to specific variables such as: Patient gender (PATGEN), patient satisfaction (PATSAT), patient ethnicity (PATETH), and physician status. Multivariate outliers, and the exploration of normality among the variables are also paramount to this statistical endeavor. The added dimension of this assignment requiring research questions formulations, and hypotheses setting are important, and warrants that the statistical analysis should provide a platform to assert: correlation among variables, and multivariate effects on specific ones. In consideration for the above reasoning, the writer/researcher can look into MANOVA to broaden the framework of analysis by looking at multiple variables.
To highlight the platform for statistical analysis, the writer/researcher proposes the following 6 research questions with their corresponding hypotheses:
RQ0: Is there a difference between patients satisfactions across physicians?
Hnull: There is no statistical significant difference between patients satisfaction across physicians categories.
Halt: There is a statistical significant difference between patients satisfaction across physicians categories.
RQ1: How can ABC Hospital survey results help improve patient care across gender lines?
H1null: Patients survey resultslead to ABC Hospital improvements across gender lines
H1alt: There is a statistical difference patients survey results are more likely to be associated with increased satisfaction among female patients.
RQ2: Can negative patients survey results have an impact across physicians categories?
H2null: Patients survey results are more likely to negatively have impact intern physicians.
H2alt: there is a statistical difference that patients survey results likely affect female interns performance.
RQ3:what is the significance of patients survey outcomes on ABC Hospital performance?
H3null: Patients survey outcomes are expected to increase ABC Hospital Performance
H3alt: there is statistically significant difference of Patients survey outcomes across ABC Hospital department.
RQ4:what is the significance of patients survey outcomes on ABC Hospitals ER and ER doctors performance?
H4null: Patients survey outcomes directly affect ER department
H4alt: there isno statistically significance difference between Patients survey outcomes and ER doctors.
RQ5: What is the significance of patients survey outcomes based on their ethnicity and satisfaction with ER services?
H5null: Attitudes towards ER services performance are directly related to patients ethnicity and satisfaction.
H5alt: There is no statistical significance that ER services performances are related to patients survey outcomes and ethnic lines.
An independent t test in table 3 was performed to compare the means for the residents physicians (M= 25.59, SD 3.944) with the interns (M=25.01, SD= 4.003). This comparison was found not to be statistically significant, t =3.901 evaluated with at 21622 degrees of freedom, and p >0.01. The Levenes test in table 4 is not significant, as shown with the use of equal variance not assumed row. Therefore for RQO, the Hnull: There is no statistical significant difference between patients satisfaction across physicians categories, could not be rejected. This result indicates that resident physicians were not functioning harder that the interns.
A one way ANOVA performed indicates that the F ratio helped to test the significance of the independent variables PYHSICIAN and STATUS. The mean squares computations were obtained by dividing the squares of each variance by the corresponding degrees of freedom. The F values are calculated by dividing the mean squares between groups by within groups (1.470/222) =7.834. This test result indicates that patient satisfaction is not statistically significant to physicians categories. It was congruent with research question 2 (RQ2), and the following hypotheses: H2null: Patients survey results are more likely to negatively have impact intern physicians. H2alt: there is a statistical difference that patients survey results likely affect female interns performance. About the post hoc comparison, it is important to note that the analysis could not be performed because there are fewer physicians in the group.
The writer/researcher attempted to perform the chi-square test and the two-way ANOVA but was not successful due to issues pertaining to an old version of SPSS, which did not generate the statistics. Unfortunately, the writer/researcher could not associate the differences between the expected and detected frequencies (Frankfort-Nachmias, & Leon-Guerrero, 2006).
Table 1: Variables Names, Labels, Values, and Measures
Variable Name
Variable Label
Values
Level of
Measurement (LoM)
N (%)
95% CI of
Mean
M (SD)
STATUS
Physician Status
0= Resident
1= Intern
Nominal
1070(33.3%)
2140 (66.7)
(.68, .67)
.67 (.471)
PHYSICIAN
Physician
1=Alexander
2= Blake
3=Charlotte
4=Daphne
5=Elizabeth
6=Finn
Nominal
452(14.1%)
430(13.6%)
632(19.7%)
708(22.1%)
524(16.3%)
458(14.3%)
(3.61,3.56)
3.56(1.596)
PATID
Patient ID
None
Nominal
PATGEN
Patient Gender
0=Male
1=Female
Nominal
1675(52.2%)
1535 (47%)
(.050,.048)
.48(.500)
PATSAT
Patient Satisfaction
None
Scale
(25.06,25.20)
25.20(3.4)
SHIFT
Day/Night Shift
0=Day Shift
1=Night Shift
Nominal
PATETH
Patient Ethnicity
1=Blue
Nominal
1193(37.2%)
(1.98,1.95)
1.95(.830)
2=Green
3=Purple
Nominal
493(30.9%)
1024(31.9%)
Total
3210(100%)
Table 2 Variables and Statistical Analyses
Research Question
Hypotheses
IV Independent
Variable(s) LoM
DV Dependent Variable LoM
Statistical Analysis Test
RQ0: Is there a difference between patients satisfactions across physicians?
Hnull: There is no statistical significant difference between patients satisfaction across physicians categories.
Halt: There is a statistical significant difference between patients satisfaction across physicians categories
PHYSICIAN
Dichotomous
PATSAT
Continuous
Independent-Samples t-test
RQ1: How can ABC Hospital survey results help improve patient care across gender lines?
H1null: Patients survey results lead to ABC Hospital improvements across gender lines
H1alt: there is a statistical difference patients survey results are more likely to be associated with increased satisfaction among female patients
Dichotommous
Scale
Independent-Samples t-test
RQ2: Can negative patients survey results have an impact across physicians categories?
H2null:Patients survey results are more likely to negatively have an impact intern physicians
H2alt: there is a statistical difference that patients survey results likely impact female interns performance
Continuous
Scale
One way ANOVA + post hoc
RQ3:what is the significance of Patients survey outcomes on ABC Hospital performance?
H3null: Patients survey outcomes are expected to increase ABC Hospital Performance
H3alt: there is statistically significant difference of Patients survey outcomes across ABC Hospital department
Continuous
Nominal
One way ANOVA + post hoc
RQ4: what is the significance of patients survey outcomes on ABC Hospitals ER and ER doctors performance?
H4null: Patients survey outcomes directly affect ER department
H4alt: there isno statistically significance difference between Patients survey outcomes and ER doctors
Continuous
Nominal
Two way ANOVA
RQ5: What is the significance of patients survey outcomes based on their ethnicity and satisfaction with ER services?
H5null: Attitudes towards ER services performance are directly related to patients ethnicity and satisfaction.
H5alt: There is no statistical significance that ER services performances are related to patients survey outcomes and ethnic lines.
Nominal
Nominal
Chi Square test of Independence
Table 3 T- test Group Statistics
Physician Status
N
Mean
Std. Deviation
Std. Error Mean
Patient Satisfaction Survey
Resident
1070
25.59
3.944
.121
Intern
2140
25.01
4.003
.087
Table 4 Independent Samples Test
Levenes Test for Equality of Variances
t-test for Equality of Means
F
Sig.
t
df
Sig. (2-tailed)
Mean Difference
Std. Error Difference
95% Confidence Interval of the Difference
Lower
Upper
Patient Satisfaction Survey
Equal variances assumed
.112
.738
3.882
3208
.000
.579
.149
.287
.871
Equal variances not assumed
3.901
2166.741
.000
.579
.148
.288
.870
Table 5 One Way ANOVA
ANOVA
Sum of Squares
df
Mean Square
F
Sig.
Physician Status
Between Groups
1.740
1
1.740
7.845
.005
Within Groups
711.593
3208
.222
Total
713.333
3209
Physician
Between Groups
15.932
1
15.932
6.267
.012
Within Groups
8155.906
3208
2.542
Total
8171.838
3209
Patient Satisfaction Survey
Between Groups
3.035
1
3.035
.190
.663
Within Groups
51147.940
3208
15.944
Total
51150.975
3209
References
Frankfort-Nachmias, C., & Leon-Guerrero, A. (2006). Social statistics for a diverse society, (4ed.). Thousand Oaks, CA: Pine Forge Press.
Meyers, L. S., Gamst, G., & Guarino, A.J. (2006). Applied multivariate research. Design and interpretation. Thousand Oaks, CA: Sage.
Get Help With Your Essay
If you need assistance with writing your essay, our professional essay writing service is here to help!
Find out more
