SEARCH · Search PubMed

Results for “Linear Regression”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 559 records · Page 31Linked to original sources

PubMed1997

The discovery of indicator variables for QSAR using inductive logic programming.

A central problem in forming accurate regression equations in QSAR studies is the selection of appropriate descriptors for the compounds under study. We describe a novel procedure for using inductive logic programming (ILP) to discover new indicator variables (attributes) for QSAR problems, and show that these improve the accuracy of the derived regression equations. ILP techniques have previously been shown to work well on drug design problems where there is a large structural component or where clear comprehensible rules are required. However, ILP techniques have had the disadvantage of only being able to make qualitative predictions (e.g. active, inactive) and not to predict real numbers (regression). We unify ILP and linear regression techniques to give a QSAR method that has the strength of ILP at describing steric structure, with the familiarity and power of linear regression. We evaluated the utility of this new QSAR technique by examining the prediction of biological activity with and without the addition of new structural indicator variables formed by ILP. In three out of five datasets examined the addition of ILP variables produced statistically better results (P < 0.01) over the original description. The new ILP variables did not increase the overall complexity of the derived QSAR equations and added insight into possible mechanisms of action. We conclude that ILP can aid in the process of drug design.

Animals↗

PubMedDate not supplied

Prediction of Henry's law constants by a quantitative structure property relationship and neural networks.

Multiple linear regression analysis and neural networks were employed to develop predictive models for Henry's law constants (HLCs) for organic compounds of environmental concern in pure water at 25 degrees C, using a set of quantitative structure property relationship (QSPR)-based descriptors to encode various molecular structural features. Two estimation models were developed from a set of 303 compounds using 10 and 12 descriptors, one of these models using two descriptors to account for hydrogen-bonding characteristics explicitly; these were validated subsequently on an external set of 54 compounds. For each model, a linear regression and neural network version was prepared. The standard errors of the linear regression models for the training data set were 0.262 and 0.488 log(H(cc)) units, while those of the neural network analogues were lower at 0.202 and 0.224, respectively; the linear regression models explained 98.3% and 94.3% of the variance in the development data, respectively, the neural network models giving similar quality results of 99% and 98.3%, respectively. The various descriptors used describe connectivity, charge distribution, charged surface area, hydrogen-bonding characteristics, and group influences on HLC values.

Journal Article↗

PubMed2005

Urinary sevoflurane and hexafluoro-isopropanol as biomarkers of low-level occupational exposure to sevoflurane.

OBJECTIVES: Sevoflurane is an inhalation halogenated anaesthetic widely used in day and paediatric surgery. We were interested in evaluating biological markers of exposure to sevoflurane, which should improve the health surveillance of occupationally exposed personnel. METHODS: A group of 36 subjects (13 male, 23 female) occupationally exposed to volatile anaesthetics in paediatric operating rooms was studied in a 2-week survey. Post-shift urine samples and specimens from passive samplers (for personal monitoring) were collected after 1.75-6 h morning exposure and analysed by headspace gas chromatography-mass spectrometry (GC-MS). Multiple determinations were assumed as independent values (in total, n = 78: 24 from men, 54 from women; 25 from smokers, 53 from non-smokers). RESULTS: Median sevoflurane external values were 0.13 parts per million (ppm) (range 0.03-18.82) (n = 78), urinary sevoflurane 0.6 microg/l urine (ND-18.5)(n = 76) and total urinary hexafluoro-isopropanol (HFIP) 0.49 mg/l urine (ND-6833.4) (n = 75). A lower limit of detection (LOD) was achieved for urinary sevoflurane (0.03 microg/l urine), allowing quantitation of all but one of the samples; >25% of urine samples were unquantifiable by HFIP and were assigned a value equal to half the LOD of 0.10 mg/l(urine). Urinary sevoflurane correlated well with breathing-zone data (r2 = 0.697 at log-log linear regression), whereas total urinary HFIP (r2 = 0.562 at log-log linear regression) seemed to be better described by a three-parameter logistic function and appeared to be influenced by smoking habits. Biological indices corresponding to National Institute for Occupational Safety and Health (NIOSH) exposure limits, calculated as means of linear regression slope and y intercept, were 3.9 mug/l(urine) and 1.4 microg/l urine for sevoflurane (corresponding to 2 ppm and 0.5 ppm, respectively), and 2.66 mg/l urine and 0.82 mg/l urine for HFIP. CONCLUSIONS: On the basis of our data, urinary unmodified, sevoflurane seems to be a more sensitive and reliable biomarker of short-term exposure to sevoflurane with respect to total urinary metabolite HFIP, which appears to be influenced by physiological and/or genetic individual traits, and seems to provide an estimate of integrated exposure.

2-Propanol↗

PubMed2005

Stature estimation based on hand length and foot length.

This study was carried out to estimate the relationship between hand length, foot length and stature using multiple linear regression analyses based on a sample of male and female adult Turks residing in Adana. Measurements of hand length, foot length and stature were taken from 155 adult Turks (80 male, 75 female) aged 17-23 years. The participants were students of the Medical Faculty of Cukurova University. A multiple linear regression model was fitted to the observed data. Stature was taken as the response or dependent variable, hand length and foot length were taken as explanatory variables or regressors. All possible (simple and multiple) linear regression models for each of males, females and both genders together were tested for the best model. The multiple linear regression model for both genders together was found to be the best model with the highest values for the coefficients of determination R2 = 0.861 and R2adjusted = 0.859, and multiple correlation coefficient R = 0.928.

Adolescent↗

PubMed1996

Occlusion maneuver to detect the relative contribution of the rib cage and abdomen to tidal volume using respiratory inductive plethysmography in infants.

Respiratory inductive plethysmography (RIP) records movements of the rib cage (RC) and abdomen (AB). A calibration procedure is needed to determine their relative contribution to tidal volume. We evaluated the hypothesis that the relative contribution of the RC and AB could be defined from respiratory efforts made during a brief occlusion of the airways in 10 infants aged 6.5-19 months, who were studied in quiet sleep. Six occlusions were performed during tidal breathing, with and without a pneumotachograph (PNT) and face mask in place. We analyzed the periods of occluded respiratory effort when RC and AB were in opposite directions (paradoxical movements), plotting RC vs. AB and performing a least-squares linear regression to estimate the ratio of the coefficients of AB/RC. Multiple linear regression of AB and RC over tidal volume during about 100 seconds of tidal breathing provided a reference standard. A ratio of 1 means that AB and RC make equal contributions to tidal volume. The feasibility of the occlusion maneuver was poor; only 51% of occlusions with PNT and 54% without led to a paradoxical movement with a good fit (r > 0.9). The mean coefficient of variation (range) was 9.35% (3.9-15.3%) with PNT and 12.1% (2.5%-26.3%) without it. The accuracy was very poor, with the mean AB/RC value being 0.94 with occlusions and 2.39 with multiple linear regression. The mean difference was 1.45 (SD 0.80), yielding 95% confidence limits for the difference of 0.12-3.01. We concluded that, due to its very poor accuracy and feasibility, the occlusion technique is not a useful method to calibrate RIP in infancy.

Abdomen↗

PubMed1989

The appropriate use of regression equations for the estimation of segmental inertia parameters.

Linear regression equations are commonly used in conjunction with experimental data to provide linear relationships between quantities which are dimensionally distinct. In many cases theoretical relationships between such quantities are known and can be used as a basis for non-linear regression equations. This study compares linear and non-linear approaches for estimating the segmental moments of inertia from anthropometric measurements using the data of Chandler et al. [Chandler et al. (1975) Investigation of inertial properties of the human body. AMRL Technical Report 74-137, Wright Patterson Air Force Base. OH.] Right limb data were used to derive the equations while left limb data were used as a cross-validation sample to evaluate the inertia estimates calculated from the equations. For the limb segments the standard error estimates had average values of 21% for the linear equations and 13% for the non-linear equations. Data on a 10 yr-old boy was used to compare the two approaches outside the sample range. The mean percentage residuals were 286% for the linear equations and 20% for the non-linear equations. A set of non-linear equations is provided.

Arm↗

PubMed2004

Analysis of nonlinear systems to estimate intraocular lens position after cataract surgery.

PURPOSE: To compare the performance of neural networks with that of linear regression to predict the postoperative effective lens position (ELP) from preoperative biometry measurements. SETTING: Departments of Ophthalmology, Medical Cybernetics and Artificial Intelligence, and Medical Physics, Medical University of Vienna, Vienna, Austria. METHODS: The neural-network-type multilayer perceptron (MLP) and a linear regression technique were used to predict ELP. Suitable MLP models and variable input combinations were selected by extended-feature subset selection. Apart from the usual preoperative biometric variables, anterior chamber depth and lens thickness were measured with partial coherence interferometry and white-to-white measurements were used as input variables. RESULTS: Prediction of ELP could be improved from a correlation coefficient (Pearson) of 0.54 for linear regression to a coefficient of 0.68 for the MLP; however, this difference was not statistically significant. CONCLUSION: The prediction of postoperative ACD with the MLP was not significantly better than the prediction using linear regression.

Adult↗

PubMed1994

The effect of a feature of regression disturbance on the efficiency of fitting growth curves.

Growth curve parameters are usually estimated by employing non-linear regression. In the present study this method was found to be inefficient for fitting growth curves, since the magnitude of random deviations of body weight greatly increases with age (heteroskedastic regression disturbance). Simulated samples of broiler body weights at different ages were generated and the associated Gompertz growth curve parameters were estimated employing three methods. Comparison of the efficiency of these methods in fitting Gompertz growth curve under this regression disturbance were performed. The results indicate that the most efficient method to estimate growth curve parameters is "weighted non-linear regression". The efficiency of this method was found to be much higher than that of conventional non-linear regression. These findings should be taken into consideration when fitting growth curves, in general, as well as for the Gompertz equation.

Age Factors↗

PubMed1975

Nutritional studies on East African herbivores. 2. Losses of nitrogen in the faeces.

1. A series of nitrogen-balance trials was done using groups of four animals of various species of wild and domesticated ruminants using pelleted diets (Arman & Hopcraft, 1975). 2. Various herbivores were given grass or grass hays, legumes, herbs and shrubs. Food and faecal samples were analysed for N. 3. With the pelleted diets, the N content of the faecal dry matter (DM) was low for eland (Taurotragus oryx Pallas), high for sheep and cattle (Bos taurus and Bos indicus) and intermediate for the three small antelope species (hartebeest (Alcelaphus buselaphus cokei Günther), Thomson's gazelle (Gazella thomsonii Günther) and duiker (Sylvicapra grimmia L.)). With the natural doffers, similar relationships were found, together with variations associated with the type of diet. 4. Three forms of equation were used to express the relationship between faecal N and food N. One of the forms, the linear regression of g faecal N/kg DM intake v. g food N/kg DM intake, fitted the results less well than did the other two and was not used to estimate metabolic faecal N(MFN) losses9 5. MFN was calculated by extrapolation of linear regressions of g faecal N/kg faecal DM v. g N intake/kg faecal DM for the pelleted diets. The range of values was (g N/kg faecal DM): Friesians 7-6, eland 8-1, zebu cattle 11-0 and small antelope and sheep 11-5-12-69 There were significant differences (P smaller than 0-001) between species in slopes and intercepts. 6. MFN was calculated from linear regressions of g faecal N/kg faecal DM v. g food N/kg food DM for all diets. This method gave the best fit for the pelleted diets. Values for these diets were (g N/kg faecal DM): eland 8-3, cattle and sheep 9-3-11-0 and small antelope 11-6-12-3. Species differences were significant (P smaller than 0-001). With grasses, values ranged from 5-9 for non-ruminants (rabbit (domesticated), warthog (Phacohoerus aethiopicus Pallas) and hippopotamus (Hippopotamus amphibius L.)) plus eland and wildebeest (Connochaetes taurinus Burchell), to 8-4 for for the other ruminants (sheep, goat, hartebeest, gazelle, duiker, buffalo (Syncerus caffer Sparrman)), kob (Adenota kob thomasi Sclater), reedbuck (Redunca redunca Pallas) and topi (Damalisucs korrigum Ogilby). For ruminants (sheep, hartebeest, duiker and kob), excluding eland, given legumes the value was 8-6, and for sheep and duiker given herbs the value was 14-7. With shrubs, faecal-N losses were variable and sometimes high. 7. With the pelleted diets, true digestibilities of crude protein (N times 6-25) varied from 0-84 to 0-91. 8. The results are discussed in relation to the digestive physiology and feeding habits of the various species, and there is an examination of the feasibility of using linear regressions of crude protein in the diet v. N in the faecal DM for evaluating the quality of the diets selected by free-ranging East African herbivores.

Animal Feed↗

PubMed1999

Comparison of three methods for the determination of baroreflex sensitivity in conscious rats.

Baroreflex sensitivity was studied in the same group of conscious rats using vasoactive drugs (phenylephrine and sodium nitroprusside) administered by three different approaches: 1) bolus injection, 2) steady-state (blood pressure (BP) changes produced in steps), 3) ramp infusion (30 s, brief infusion). The heart rate (HR) responses were evaluated by the mean index (mean ratio of all HR changes and mean arterial pressure (MAP) changes), by linear regression and by the logistic method (maximum gain of the sigmoid curve by a logistic function). The experiments were performed on three consecutive days. Basal MAP and resting HR were similar on all days of the study. Bradycardic responses evaluated by the mean index (-1.5 +/- 0.2, -2.1 +/- 0.2 and -1.6 +/- 0.2 bpm/mmHg) and linear regression (-1.8 +/- 0.3, -1.4 +/- 0.3 and -1.7 +/- 0.2 bpm/mmHg) were similar for all three approaches used to change blood pressure. The tachycardic responses to decreases of MAP were similar when evaluated by linear regression (-3.9 +/- 0.8, -2.1 +/- 0.7 and -3.8 +/- 0.4 bpm/mmHg). However, the tachycardic mean index (-3.1 +/- 0.4, -6.6 +/- 1 and -3.6 +/- 0.5 bpm/mmHg) was higher when assessed by the steady-state method. The average gain evaluated by logistic function (-3.5 +/- 0.6, -7.6 +/- 1.3 and -3.8 +/- 0.4 bpm/mmHg) was similar to the reflex tachycardic values, but different from the bradycardic values. Since different ways to change BP may alter the afferent baroreceptor function, the MAP changes obtained during short periods of time (up to 30 s: bolus and ramp infusion) are more appropriate to prevent the acute resetting. Assessment of the baroreflex sensitivity by mean index and linear regression permits a separate analysis of gain for reflex bradycardia and reflex tachycardia. Although two values of baroreflex sensitivity cannot be evaluated by a single symmetric logistic function, this method has the advantage of better comparing the baroreflex sensitivity of animals with different basal blood pressures.

Analysis of Variance↗

PubMed1996

Genetic components for milk production of tarentaise, hereford, and tarentaise x hereford cows.

The objective of this study was to estimate genetic components for milk production of Hereford (HH), Tarentaise (TT), and Hereford x Tarentaise or Tarentaise x Hereford (HT) cows under range conditions at the Northern Agricultural Research Center, Havre, MT. Milk production of 494 cow-calf pairs was estimated by the weigh-suckle-weigh procedure from 1989 to 1994. Milk production and cow and calf weights were measured at 40 d (early) and 120 d (late) of lactation. The mating scheme was a 3 x 3 design in which HH, HT, and TT cows were mated with HH, HT, and TT sires, producing calves that were varying percentages of Hereford and Tarentaise. Data were analyzed by least squares procedures according to a model that included year, age of cow, sex of calf, linear regression on calf age, and linear regressions on coefficients that coded for various genetic components. These genetic components were breed individual, breed maternal, breed grand maternal, individual heterosis (calf heterosis), and maternal heterosis (cow heterosis). Coefficients for genetic components were such that breed effects estimated TT-HH and regressions on heterosis values estimated the full effect of heterosis. The breed individual genetic component was significant only for calf early and late weights; Tarentaise calves were lighter. Breed maternal was significant for most traits, and these effects were positive for milk production (2.16 and 3.77 kg/d for early and late, respectively) and calf weights but negative for cow weights. Individual heterosis was not significant for early milk production but was significant for late milk production (1.05 kg/d) and calf weights (2.3 and 8.1 kg for early and late, respectively). Maternal heterosis was significant for most traits and increased early milk production (1.25 kg/d), calf weights (4.5 and 5.8 kg for early and late, respectively), and cow weights (18.1 and 6.0 kg for early and late, respectively).

Analysis of Variance↗

PubMed1990

A simplified competition data analysis for radioligand specific activity determination.

Non-linear regression and two-step linear fit methods were developed to determine the actual specific activity of 125I-ovine prolactin by radioreceptor self-displacement analysis. The experimental results obtained by the different methods are superposable. The non-linear regression method is considered to be the most adequate procedure to calculate the specific activity, but if its software is not available, the other described methods are also suitable.

Animals↗

PubMed2004

High-sensitivity C-reactive protein concentration in a healthy Brazilian population.

BACKGROUND: The use of serum high-sensitivity C-reactive protein (hs-CRP) for cardiovascular risk profiling requires knowledge of the distribution in different populations. We studied serum hs-CRP in a healthy Brazilian population, with no evidence of heart disease, relative to gender and smoking status as well as other clinical and laboratory variables. METHODS: 684 healthy Brazilian individuals, aged 14-74 years (mean 40.6 years), 295 men (43.1%) and 389 women (56.9%), were enrolled between July 1998 and July 2001. The relationship between the log-transformed hs-CRP, stratified by gender and smoking status (non-smokers, smokers), and clinical and laboratory variables were analyzed using the Spearman correlation coefficient and multiple linear regression. RESULTS: In a multiple linear regression model age (beta=1.0069, p=0.03), body mass index (BMI) (beta=1.0093, p<0.01), total cholesterol/HDL cholesterol ratio (TC/HDL-C) (beta=1.1015, p<0.01) and heart rate (beta=1.0139, p<0.01) were independently correlated with hs-CRP. Stratification according to gender and smoking was able to disclose different interactions between these variables and hs-CRP. Variables significantly correlated in each stratum were as follows: in smoker men, age (beta=1.0236, p=0.04) and TC/HDL-C (beta=1.2065, p<0.01); in non-smoker men, BMI (beta=1.0786, p<0.01) and TC/HDL-C (beta=1.1397, p=0.01); in smoker women, BMI (beta=1.1006, p=0.03); and in non-smoker women, BMI (beta=1.0873, p<0.01), TC/HDL-C (beta=1.1405, p=0.01) and heart rate (beta=1.0165, p=0.02). CONCLUSIONS: The clinical and laboratory variables studied influenced the concentration of hs-CRP according to gender and smoking. In assessing the risk of cardiovascular events in relation to serum hs-CRP level, stratification by gender and smoking status is indicated.

Adolescent↗

PubMedDate not supplied

Measurement of adrenal corticosteroids by high-performance liquid chromatography.

OBJECTIVE: To develop a high-performance liquid chromatography (HPLC) method for simultaneous measurement of adrenal corticosteroids important for the diagnosis of congenital adrenal hyperplasia. DESIGN: Blind comparison using convenience samples. SETTING: Research laboratory at the Department of Medical Technology, Bowling Green State University. PATIENTS: Referred samples for cortisol, 11-deoxycortisol (S), and 17-hydroxyprogesterone (OHP). Stimulation condition unknown. INTERVENTION: Venous blood collected, processed to serum, and frozen. Adrenal steroids measured by radioimmunoassay (RIA) and by HPLC. MAIN OUTCOME MEASURES: Linearity, analytic sensitivity, accuracy, and precision were evaluated using recovery, coefficient of variation (c.v.), correlation coefficient (r), Student's t test, F test, and linear regression. RESULTS: Calibration curves were linear to at least 50 mumol/L, and the low-end sensitivity was 10 nmol/L. Analytic recovery ranged from 89.8% to 103.9%. The CV was below 7.5% (n = 10) for all three steroids. Comparison to RIA yielded r values of 0.909 (n = 25), 0.932 (n = 10), and 0.983 (n = 10) for cortisol, S, and OHP, respectively. Student's t-test results were 1.13 (p = 0.27), 0.10 (p = 0.93), and 0.48 (p = 0.64) for cortisol, S, and OHP, respectively. F-test results were 1.14 (p = 0.75), 1.37 (p = 0.66), 2.39 (p = 0.21) for cortisol, S, and OHP, respectively. Results for cortisol showed a negative bias of 97.0 nmol/L (3 mug/dL). Linear regression analysis revealed proportional errors of -15.0% -7.0%, and +9.0%, for cortisol, S, and OHP, respectively. CONCLUSION: (ABSTRACT TRUNCATED AT 250 WORDS)

Adrenal Cortex Hormones↗

PubMed2000

Plasma levels of remnant particles are determined in part by variation in the APOC3 gene insulin response element and the APOCI-APOE cluster.

Remnant particles of triglyceride-rich lipoproteins (RLP) are known to be a strong predictor of atherogenicity. The serum concentrations of remnant-like particle triglyceride (RLPTG) and remnant-like particle cholesterol (RLPC) have been determined in a representative sample of the Czech MONICA study (n = 285). The relationship was investigated between remnant particle triglyceride/cholesterol concentrations and polymorphisms in the genes APOC3 (-482C-->T/3238C-->G), APOE (epsilon2/epsilon3/epsilon4), APOCI (-317-321ins), APOB (signal peptide), hepatic lipase (LIPE, -480C-->T), and lipoprotein lipase (LPL, S447X). Univariate analysis showed significant effects on RLPTG associated only with the APOE genotype (P = 0.009), the APOC3 -482C-->T genotype (P = 0.018), and the APOCI -317-321ins (P = 0.014) genotype and significant effects on RLPC with APOE (P = 0.01) and APOCI -317-321ins (P = 0.021). The raising effect of the APOE genotype for both remnant cholesterol and triglyceride was confined to the epsilon2/4 (n = 6) and varepsilon4/4 (n = 3) groups, and thus when the epsilon2/4 group was omitted in order to analyze by allele (epsilon2+/epsilon3+/epsilon4+), significance was lost (P = 0.6). There was strong linkage disequilibrium between the APOE and APOCI alleles (chi(2), P < 0.001) and a multivariate ANOVA of RLPTG with all three significantly associated variants as factors demonstrated that while the APOC3 -482C-->T effect was independent of the others (P = 0.003), the APOCI -317-321ins and APOE effects were not. This was also true for the APOCI -317-321ins and APOE effects on RLPC. To assess whether APOE-CI effects on RLPC were independent of their effects on total cholesterol and triglyceride levels, multiple linear regression was used. Using multiple linear regression, it appeared that the APOE-CI effects on RLPC were independent of their effects on plasma cholesterol, but the effects of APOC3 and APOE-CI on RLPTG could not be separated from their effects on plasma Tg levels. Further characterization of this remnant particle phenotype and its genetic determinants may lead to a better understanding of its metabolism and contribution to atherosclerosis.

Adult↗

PubMed1986

Offspring-parent genotypic regression: how linear is it?

Linearity of offspring-parent regression is investigated for quantitative characters with dominance. This is done by fitting a second-order regression equation in orthogonal bivariate polynomials to the offspring's genotypic value and evaluating the significance of nonlinear terms in the equation. Three genetic models are considered.

Animals↗

PubMed1975

[A new method for calculating a compensation function in power series].

A new method for the derivation of a compensation function in power series is developed and compared with the linear regression of the logarithms. At the same time from the formulas of the linear regression corresponding formulas for the regression of power series without the need of logarithms are derived. Comparing the newly developed method of the "persistence report" with the linear regression of the logarithms it results that the new method often reaches a better approach of the values of the compensation function to the measured values; furthermore the new formulas are simpler and applicable without computers.

Biometry↗

PubMedDate not supplied

Applicational possibilities of linear and non-linear (polynomial) regressions and analysis of variance: kinetics of chloramphenicol.

The dependance of the content and microbiological activity of Chloramphenicol (active substance) at dissolution on time and on pH, as well as that of the content and microbiological activity of Chloramphenicol (250 mg capsules) at release, on time, in in-vitro conditions, was determined using linear and non-linear (polynomial and generalized dilution) regressions. Based on the square error value, the dependance of the content and microbiological activity of Chloramphenicol (active substance) at dissolution on time and different pH values, as well as the dependance of the content and microbiological activity of Chloramphenicol (capsules) at release on time were best described by polynomial function. The comparison of the content and microbiological activity of Chloramphenicol (active substance) at dissolution at different pH values, as well as of Chloramphenicol (capsules) at release showed the significant correlation between these parameters (r = 0.999, P << 0.001). The comparision of the content, on one hand, and microbiological activity of Chloramphenicol (active substance), on the other, at dissolution at different pH values, as a function of time, was done using a modified method of one-way analysis of variance for linear regression comparisons. Based on the value of Fischer's coefficient (F), there is a statistically very significant difference between the contents and between the microbiological activities of chloramphenicol (active substance) at dissolution and different pH as a function of time (P << 0.005).

Algorithms↗