The Chi Square Test Health And Social Care Essay

The Chi Square Test Health And Social Care Essay

1.a. The statistical trial that would be run to look into whether the difference in attending rates between the different classs of patients was statistically important is the Chi-square trial.

Chi-square is appropriate because it assumes the homogeneousness of informations distribution between categorical variables, and where differences in such distributions arise, it tests the significance. Chi-square assumes that variables are independent ( its nothing hypothesis ) , and that informations are indiscriminately collated instead than being assigned from a fixed population as in this instance. The appreciably big sample size is besides a warrant for the truth of Chi-square trial.

1.b. At the clip of directing the invitations, the GPs had expected equal figure of patients to go to for each of the classs, therefore taking to the void hypothesis ( premise of independency ) that there is no important difference in the rate of attending for each of the groups represented in the survey. However, the observations were tested between the 2678 invited patients, who were grouped based on their Sexual activity ( Males and Females ) , Age ( 35-44 old ages, 45-54 old ages, and 55-64 old ages ) , Smoking position ( Smokers and Non-smokers ) , Body weight ( Obesed and Not obesed ) , and Diet Scores ( 1,2,3, and 4 for Best, Good, Bad, Worst scores severally ) . The consequences are shown on the affiliated PASW end product ( Tables 1.1-1.5, and Fig. 1.1-1.5 ) .

Table 1.1a shows the crosstabulation end product of the attending in males and females. Harmonizing to the tabular array, males ( 987 ) were less likely to go to than females ( 1218 ) , and were every bit good more likely non to go to ( 262 ) , than the females ( 211 ) . Chi-square trial ( Table 1.1b ) confirm that there is a important ( X2=17.68, df=1, Fisher ‘s P & A ; lt ; 0.05 ) .

In Table 1.2a, attending tend to change between the age groups in the order of 812, 732, and 661 for those with 35-44, 45-54, and 55-64 old ages groups. However, there is no association ( X2=2.074, df=2, P & A ; gt ; 0.05 ) between such variableness in attending to the age differences.

Table 1.3a-b shows the response of tobacco users and non-smokers to invitation. More non-smokers ( 1583 ) than expected were in attending, while lesser figure of tobacco users ( 622 ) attended. On the other manus, more tobacco users ( 208 ) were likely non to go to than non-smokers ( 265 ) , giving the standardized residuary value of 5.1 and -3.4 severally. A significantly higher ( X2=45.263, df=1, P & A ; lt ; 0.05 ) attending rate was observed for the non-smokers compared the tobacco users. This is consequence is besides supported by really high likeliness and linear-by-linear association as shown in Table 1.3b.

Table 1.4a-b is the crosstabulation of the association of attending with organic structure weight. The consequence shows that organic structure weight determined that rate of attending of the patients. Therefore, the obesed patients were less ( X2=10.2, df=1, P & A ; lt ; 0.05 ) probably to go to ( 197 ) compared with the ‘not obesed ‘ patients ( 2008 ) .

Table 1.5a-b compared the attending rate of patients holding different dietetic tonss. The result revealed a likeliness of attending in the order of 1 ( 632 ) & A ; gt ; 3 ( 564 ) & A ; gt ; 2 ( 541 ) , and 4 ( 468 ) . This order of attending by patients of different dietetic mark was found to be important ( X2=38.259, df=3, P & A ; lt ; 0.05 ) .

1.c. Sexual activity, smoking position, organic structure weight and diet mark affect the rate of attending of patients in response to invitation by GPs in Manchester. Thus, on the footing of sex, more females are likely to react to an invitation than males. Besides, more non-smokers and ‘not obesed ‘ are likely to go to than tobacco users and obesed severally. On the footing of diet tonss, patients with diet mark of 1 were more likely to go to than those of mark 3, who were every bit good more likely to go to than those on 2. the latter group are more likely to go to than those on 4.

A critical expression of the result of this survey could uncover some implicit in grounds for the important consequences. It could be that males, tobacco users, and obesed were afraid of returning to the infirmary as they think a wellness cheque could uncover some upseting medical conditions or that they were non satisfied with the quality of wellness attention services they had received. For those who attended, it might be that they are funny about cognizing to what extent their conditions would hold improved.

The ascertained differences in the rates of attending between the age groups were statistically non important. This is to state that the GPs expected the figure of patients who attended for each age group to go to, while besides being cognizant that the figure of patients who did non go to, would besides non hold attended. Therefore, whether or non patients responded to the invitation is independent ( non influenced ) by their age groups.

In drumhead, these findings could pave manner for farther probes into unknoting some of the grounds why those who attended did so and frailty versa.

1.d. My reading of the informations will be that the void hypothesis ( premise of independency ) merely holds true if the age groups of the patients is to be considered when ask foring them, otherwise, the rate of attending or response to invitation is mostly associated with the sex, smoking position, organic structure weight, and diet mark of the patients.

1.e. The consequence of this survey will assist the GPs in Manchester in planning ; given them an penetration on which of the classs of the patient ‘s population that should have more attending during subsequent invitations. For case, since the patient ‘s response to invitation is independent on the age groups, the GPs may make up one’s mind non to see that class in subsequent invitation since there is non likely to do any difference.

On the other manus, if an invitation is to be sent based on their sex, organic structure weight, smoke, and diet mark differences, the GPs would hold to take excess steps, including reexamining their services and stepping up enlightenment, if they are to guarantee that every bit many as are invited would go to.


One-way analysis of discrepancy ( ANOVA ) will be most suited for finding whether increased nanotube concentration causes increased tensile strength, This is because, ANOVA tests for the grade of variableness between the mean of quantitative informations ( herein represented by the Tensile strength ) that were independently and indiscriminately sampled from group variables ( samples treated with 5 % , 10 % , 15 % , and 20 % nanotube concentrations ) . The premise is that there is homogeneousness of discrepancies and that increasing nanotube concentration does non ( void hypothesis ) have any consequence on tensile strength.

Some steps ( see attached PASW Output Tables 2.1-2 ) have been taken to guarantee that these informations do conform with the premises of ANOVA, hence its suitableness.

Normality: appraisal of the symmetricalness ( lopsidedness ) and its bunch around the mean ( kurtosis ) was done foremost by utilizing the values in the descriptive end product. As a regulation of the pollex, divergence from normalcy can be assumed if the lopsidedness is by far larger than its standard mistake. Here both of the lopsidedness ( -0.223 ) and kurtosis ( -0.476 ) , though negative, are good near to zero and rather smaller than their standard mistake of 0.472 and 0.918 severally. Kolmogorov-Smirnov and Shapiro-Wilk trials ( Table 2.3 ) both uphold the premise that the informations are usually distributed, and are supported by the mentality of the histogram, Normal Q-Q secret plans, and box secret plans.

The premise of homogeneousness of the discrepancies was every bit good confirmed utilizing the Levene statistics, which accepts the void hypothesis that the discrepancies are equal ( P & A ; gt ; 0.05 ) . Looking at the standard divergence and standard mistake for the several group, it could be seen that there is an overall lessening of fluctuation in tensile strength as the nanotube concentration increased. This is besides supported by the mean secret plan ( Fig. 2.4 ) .

Finally, to look into farther for this suggested influence ( see Fig. 2.4 ) , the F-test on the ANOVA tabular array is used. The consequence shows that there is a important ( P & A ; lt ; 0.05 ) difference between the mean tensile strength of the different samples treated at different nanotube concentrations. This implies that the obtained F-test is higher than the critical value in F-distribution tabular array

Because ANOVA can merely place differences in sample mean between groups without foregrounding where precisely the differences lie, it is imperative that the Post-hoc trial ( e.g. Least important difference as in this survey ) be used to observe which of the group means differ. The Post-hoc LSD was discretionarily chosen and it revealed that though the average difference between samples treated with 10 % and 15 % nanotube concentrations were non important ( P & A ; gt ; 005 ) , other group means differed ( P & A ; lt ; 0.05 ) between each other and even with 10 % and 15 % samples.

Decision: Increasing nanotube concentration gives a important addition in tensile strength.


The hypothesis tested here is that the drug Captopril has a important consequence ( may be higher or lower ) on the systolic and diastolic force per unit area in the reasonably hypertensive topics.

The corresponding void hypothesis is that Captopril does non hold any consequence on the systolic or diastolic blood force per unit area degree. Therefore, the void hypothesis suggests that the average difference of the ‘before ‘ and ‘after ‘ intervention systolic and diastolic blood force per unit area will be severally zero.

The ideal statistical method here is paired T-test. Paired T-test enables the comparing of differences ( if any ) in population agencies of observations made for each instance both before and after they are being subjected to a trial. The trial aims to understate, and if possible eliminate external mistakes or interventions that could take to false positive ( type I error ) or false negative ( type II mistake ) consequences.


Independence: This survey involves two wholly separate groups ( systolic and diastolic blood force per unit area group ) . It is worthy to foreground that both groups are non in any manner related to each other, therefore alterations in any one of them is non expected to act upon the other. Besides, the premise of independency is supported by the fact that there is no interaction between the before and after agencies as the latter is merely expected to be due to the consequence of the drug.

Normal distribution: Premise of normal distribution and equal discrepancy is confirmed in the descriptive statistics, every bit good as the box secret plans. All instances of ‘before ‘ and ‘after ‘ intervention for the two classs, have really little lopsidedness and kurtosis, with the latter falling a spot outside -1 for the ‘before ‘ systolic diastolic ( see Table 3.1 ) . However, the Normal Q-Q secret plans for each of the instance shows some normal distribution, which is besides supported by the box secret plans. The two outliers on the opposite sides of the ‘after -treatment ‘ diastolic blood force per unit area has been ignored, giving that the lopsidedness ( 0.191 ) and kurtosis ( -0.341 ) are within acceptable scopes ( See Table 3.2 ) , and that Kolmogorov-Smirnov and Shapiro-Wilk are both continuing that the instances in the before and after groups are usually distributed. It is worthy to advert that attending is non given to prove for homogeneousness of discrepancy as Paired t-test is non really rigorous on equality and non-equality of discrepancies.

Testing for the consequence of Capoten on blood force per unit areas

The result of trial is attached as PASW end product ( Tables 3.4-3.6 ) . Table 3.4 displays the mean, standard divergence, and standard mistake for each of the classs. Ocular analysis shows that intervention with Capoten caused a bead in systolic blood force per unit area to 158.20+20 mmHg from 176.93+20.57 mmHg, while diastolic blood force per unit area besides reduced to 103.07+12.55 mmHg from 112.33+10.47 mmHg. However, it remains unknown, if this bead is statistically meaningful or non.

The mated sample trial ( Table 3.5 ) shows that non merely did the systolic and diastolic force per unit areas bead by 18.93 mmHg, and 9.26 mmHg severally, the autumn was really important ( P & A ; lt ; 0.05 ) . It besides shows that merely a standard divergence of up to 9.02 mmHg and 8.61 mmHg, every bit good as the assurance intervals of 13.93-23.93 mmHg and 4.5-14.04 mmHg for the average difference in systolic and diastolic severally. It is merely within this limits that subsequent trial consequences to be assumed to be accurate.

Finally, a correlativity ( see Table 3.6 ) of the mated agencies revealed an about absolutely consistent and unvarying autumn in both systolic ( r=0.901, P & A ; lt ; 0.05 ) and diastolic ( r=0.734, P & A ; lt ; 0.05 ) blood force per unit area, after the topics were treated with Capoten.

Decision: The consequence of this survey solidly rejects the void hypothesis ( statement of no consequence ) , and accepts the alternate hypothesis which states that Capoten affects blood force per unit area. In these braces, the consequence is a important decrease in systolic and diastolic blood force per unit area after disposal of Capoten.


Having noted the qualitative variable ( sexual activity ) that should move as an account for response variable ( longevity/lifespan ) , a one-way ANOVA was used to research the extent of association between the two variables. Sexual activity was used to group the males ( n=125 ) into 25 per group of those life:


1 receptive female ( RF )

8 receptive females

1 non-receptive female ( NRF )

8 non-receptive females.

a, vitamin D, and vitamin E are controls.

Normality of length of service distribution was assumed, given the visual aspect of the histogram, and supported by the lopsidedness and kurtosis as shown in Table 4a. II. Levene trial for homogeneousness of discrepancies confirmed the premise of equal discrepancy.

The consequence from ANOVA shows that there is important difference ( p & A ; lt ; 0.05 ) in the average length of service of the males as grouped. The mean secret plan shows a diminution in the lifetime of males who live with either 1 or 8 receptive females. Therefore, a multiple comparing was made to happen out whether the ascertained diminution was important. The result revealed that the lifetime of the controls ( 63.56, 64.80, and 63.36 yearss for males populating entirely, with 1 NRF, and 8 NRF severally ) , were significantly higher ( p & A ; lt ; 0.05 ) than that of those populating with either 1 ( 56.76 yearss ) or 8 ( 38.72 yearss ) receptive females. There was besides an ascertained difference ( p & A ; lt ; 0.05 ) between the length of service of males populating with 1 receptor and those populating with 8 receptors.

The findings in this survey suggest that the sexual activities of male fruitflies has impact on their length of service, such that the more receptive females they live with, the lesser their lifetime.

4b. To hold a broader position of whether sleep and/or organic structure size have impacts on lifetime, multiple linear ( enter ) arrested development was used. Normality of the distribution was confirmed utilizing histogram and Normal P-P secret plans of Regression Standardized residuary, while the spread secret plan showed that the distribution is random. Furthermore, matrix spread diagram appears to propose a additive association, particularly with length of thorax ) implying that a additive relationship exist between any of the variables and length of service. The theoretical account has a singular correlativity ( R=0.638 ) with a little criterion mistake of estimation ( 13.641 ) , and the F ( 41.787 ) for the arrested development is important ( p & A ; lt ; 0.05 ) , therefore corroborating that the theoretical account is tolerably additive. The coefficient of correlativity shows that there is important correlativity ( r=0.636, P & A ; lt ; 0.01 between the lifetime and length of thorax, while no important correlativity exist between lifetimes and yearss spent kiping.

In decision, lessening in length of service is associated with addition in the figure of receptive female life with a male fruitfly. On the other manus, length of service of male fruitflies increases with an addition in their thorax length, but is non affected by the figure of yearss they spend kiping.