Differences

This shows you the differences between two versions of the page.

--- principles:low_coupling [2021-10-18 21:49] – +++ restored +++ christian
+++ principles:low_coupling [2025-11-10 04:32] (current) – created mellissam44
@@ Line 1: / Line 1: @@
-====== Low Coupling (LC) ======
+How to Calculate P Values: A Clear and Confident Guide
+Calculating p-values is an important part of statistical analysis that helps researchers determine the significance of their results. A p-value is a measure of the probability that an observed difference between groups is due to chance, rather than a true difference in the population. In other words, it tells us whether our results are statistically significant or not.
-===== Variants and Alternative Names =====
-  * Loose Coupling
-===== Context =====
+To calculate a p-value, researchers need to perform a hypothesis test. This involves comparing the observed data to what would be expected if there were no real difference between groups. If the observed data is significantly different from what would be expected by chance, then the p-value will be low, indicating that the results are statistically significant. On the other hand, if the observed data is not significantly different from what would be expected by chance, then the p-value will be high, indicating that the results are not statistically significant.
-/* fill in contexts here: */
+(Image: [[https://calculator.city/wp-content/uploads/2024/07/Designer-4.jpeg|https://calculator.city/wp-content/uploads/2024/07/Designer-4.jpeg]])Understanding P Values
-  * [[contexts:Object-Oriented Design]]
+[[https://www.youtube.com/embed/JQc3yx0-Q9E|external frame]]
-  * [[contexts:API Design]]
+Definition of P Values
-  * [[contexts:Architecture]]
+P value is a statistical measure that represents the probability of obtaining a result as extreme as, or more extreme than, the observed result under the null hypothesis. The null hypothesis is the default assumption that there is no significant difference between two groups or variables. The P value is used to determine whether the null hypothesis should be rejected or not.
-===== Principle Statement =====
-[[glossary:Coupling]] between [[glossary:modules]] should be low.
+A P value ranges from 0 to 1, with a smaller P value indicating stronger evidence against the null hypothesis. In general, a P value of 0.05 or less is considered statistically significant, which means that the result is unlikely to have occurred by chance alone.
+Significance in Hypothesis Testing
-===== Description =====
+P values play a crucial role in hypothesis testing, which is a method used to determine whether there is enough evidence to reject the null hypothesis. In hypothesis testing, researchers formulate a null hypothesis and an alternative hypothesis, and then collect data to test the hypotheses.
-A module should not interact with too many other modules. Furthermore if a module //A// interacts with another module //B//, this interaction should be loose, which means that //A// should not make too many assumptions about //B//.
-Coupling is a measure of dependency between modules. The more dependencies there are, the stronger the dependencies are, and the more assumptions are made upon other modules, the higher is the coupling.
+If the P value is less than the significance level (usually set at 0.05), the null hypothesis is rejected, and the alternative hypothesis is accepted. This means that there is strong evidence to support the alternative hypothesis and that the result is not due to chance.
-There are different forms of couplings which can be rated according to their strength((G. J. Myers: //Reliable Software through Composite Design//)):
+P Values in Context
-  * //No coupling//: The modules do not know each other.
+It is important to note that P values alone cannot determine the validity of a study or the importance of a result. P values should always be interpreted in the context of the study design, sample size, and other relevant factors.
-  * //Call coupling//: A module calls another one.
-  * //Data coupling//: A module calls another one passing parameters to it.
-  * //Stamp coupling//: A module calls another one passing complex parameters to it.
-  * //Control coupling//: A module influences the control flow of another module.
-  * //External coupling//: The modules communicate using a simple global variable.
-  * //Common coupling//: The modules communicate using a common global data structure.
-  * //Content coupling//: A modules depends on the inner working of another module. This is the strongest form of coupling.
-The forms ranging from no coupling to stamp coupling can be considered "good" couplings. The others are rather strong.
-There are also some additional forms of undesirable couplings:
+Additionally, P values should not be used as the sole criterion for decision-making in scientific research. It is essential to consider other factors, such as effect size, confidence intervals, and clinical relevance, when interpreting research findings.
-  * //Tramp coupling//: A module is only coupled to a data structure because some other module needs the data. The module gets the data and passes it to the other module without touching the "tramp data" ((M. Page-Jones: //The Practical Guide to Structured Systems Design//)).
-  * //Logical coupling//: A module makes some assumptions about another module without referencing it. For example a module //A// only sorts a list because some other module //B// which //A// technically does not know about needs it sorted.
+In summary, P values are a statistical measure used to determine the probability of obtaining a result as extreme as, or more extreme than, the observed result under the null hypothesis. They play a critical role in hypothesis testing, but should always be interpreted in the context of the study design and other relevant factors.
+Calculating P Values
-===== Rationale =====
-If a module //A// interacts with a module //B//, there is a certain dependency between these modules. When for example //A// uses a certain functionality of //B//, then //A// depends on //B//. //A// makes the assumption that //B// provides a certain service, and moreover it makes assumptions on how this service can be used (by which mechanism, which parameters, etc.). If one of these assumptions is not true anymore because //B// has changed for some reason, //A// also has to change. So the fewer dependencies there are, the less likely it is that //A// stops working and has to be changed.
+Calculating P values is an essential step in hypothesis testing. It helps determine the probability of obtaining a sample mean as extreme as the one observed, assuming the null hypothesis is true. Here are the steps involved in calculating P values:
-Furthermore if //A// makes many and detailed assumptions about //B//, there is also a high probability that //A// has to change despite only relying on one other module. This is because in such a case //A// also needs to change when only a certain detail of //B// changes.
+Preparation of Data
-But if coupling is low, there are only few assumptions between the modules which can be violated. This reduces the chance of [[glossary:ripple effects]].
+To calculate P values, you need to have a dataset that meets certain assumptions. The data should be normally distributed, and the sample size should be large enough to satisfy the central limit theorem. Additionally, the data should be independent and randomly sampled from the population.
+Selection of Statistical Test
-===== Strategies =====
+The selection of the appropriate statistical test depends on the type of data you have and the research question you are trying to answer. For example, if you have two independent groups, you can use a two-sample t-test, while a one-sample t-test can be used to compare the mean of a sample to a known value. If you have more than two groups, you can use an ANOVA test.
-  * Indirection: Don't access the other module directly but have another module do that.
+Execution of the Test
-  * Dependency Inversion/Abstract Couplings:
-  * Use lower form of coupling
-  * Merge modules: when there is only one module, then there is no communication and thus no coupling
-  * Hide information: Information which is hidden cannot be depended upon.
-===== Caveats =====
+Once you have selected the appropriate statistical test, you can execute the test using software such as R or SPSS. The software will calculate the test statistic, which is used to determine the P value. The test statistic is calculated by comparing the observed sample mean to the expected population mean under the null hypothesis.
-Coupling can be reduced by several technical measures (see [[#strategies]]). But while these measures reduce the coupling technically, they do not necessarily reduce the logical coupling. In such a case two modules A and B may seem decoupled, but ripple effects may occur anyway because of the logical coupling. In such a case it is better to make the coupling explicit by not applying a decoupling strategy. It may also be possible to find a better suitable strategy or a better way of applying the strategy to also get rid of the logical coupling.
+Interpreting Test Results
-Furthermore note that coupling to a stable module is often no problem. The problematic cases are couplings to unstable modules. This means that applying decoupling strategies is beneficial when a coupling to an unstable module is reduced. But it may not be beneficial in the other cases.
+After executing the test, you will obtain a P value. If the P value is less than the significance level (usually 0.05), then you can reject the null hypothesis and conclude that there is sufficient evidence to support the alternative hypothesis. If the P value is greater than the significance level, then you fail to reject the null hypothesis and conclude that there is insufficient evidence to support the alternative hypothesis.
-See also section [[#contrary principles]].
+In conclusion, calculating P values is an important step in hypothesis testing. By following the steps outlined above, you can accurately determine the probability of obtaining a sample mean as extreme as the one observed, assuming the null hypothesis is true.
+P Values and Sample Size
-===== Origin =====
+Effect of Sample Size on P Values
-/* the *primary* source */
-  * W. P. Stevens,  G. J. Myers, L. L. Constantine: //Structured design//
-===== Evidence =====
+Sample size is an important factor in determining the accuracy of statistical tests. The larger the sample size, the more reliable the results of the test. When the sample size is small, the p-value can be misleading. A small p-value may indicate statistical significance, but it may not be practically significant. On the other hand, a large sample size can make even small differences statistically significant.
-/* Comment out what is not applicable and explain the rest: */
-/*  * [[wiki:Proposed]]*/
-  * [[wiki:Examined]]: There are metrics that try to measure coupling and there are studies relating these coupling measures to the number of errors found during testing (({{page>resources:A Handbook Of Software And Systems Engineering#reference}})). This correlation is evident. The limitation of these studies is that these coupling metrics cannot represent the coupling notion completely.
-  * [[wiki:Accepted]]: The concept of low coupling is widely known and described in several well-known books for example in {{page>resources:Applying UML And Patterns#reference}}.
-/*  * [[wiki:Questioned]]*/
+For example, if a study with a small sample size finds a p-value of 0.05, it may indicate that there is a significant difference between the groups being compared. However, if the same study were repeated with a larger sample size, the p-value may be much smaller, indicating an even more significant difference.
+Adjusting for Sample Size
-===== Relations to Other Principles =====
+To adjust for the effect of sample size on p-values, researchers often use a correction factor called the Bonferroni correction. This correction factor adjusts the p-value threshold for multiple comparisons. For example, if a study is comparing three groups, the p-value threshold for statistical significance should be adjusted to 0.017 (0.05/3) to avoid false positives.
-==== Generalizations ====
-==== Specializations ====
+Another way to adjust for sample size is to use effect size measures such as Cohen's d or Hedges' g. These measures take into account the size of the difference between the groups being compared, as well as the sample size. By using effect size measures, researchers can determine the practical significance of the results, in addition to the statistical significance.
-  * [[Constantine's Law]]: Constantine's Law is just the combination of the two principles LC and HC.
-  * [[Dependency Inversion Principle]] (DIP): LC aims at reducing the dependencies to other modules. One way to do so is to only depend on abstractions. DIP is about this aspect.
-==== Contrary Principles ====
+In summary, sample size plays an important role in determining the accuracy and reliability of statistical tests. A larger sample size can make small differences statistically significant, while a small sample size may produce misleading p-values. Researchers can adjust for the effect of sample size by using correction factors such as the Bonferroni correction, or by using effect size measures such as Cohen's d or Hedges' g.
+Types of Errors
-  * [[Keep It Simple Stupid]] (KISS): Reducing the coupling often involves the use of complicated interaction patterns, indirections, etc.
+Type I and Type II Errors
-  * [[High Cohesion]] (HC): A system consisting of one single module has a very low coupling as there are no dependencies on other modules. But such a system also has low cohesion. The other extreme, very many highly cohesive modules, naturally has a higher coupling between the modules. So here a compromise has to be found.
-  * [[Rule of Explicitness]] (RoE): Direct communication typically has the disadvantage of a higher coupling. Indirection reduces coupling but creates implicit/indirect communication paths.
-==== Complementary Principles ====
-  * [[Tell, don't Ask/Information Expert]] (TdA/IE): IE may help to reduce coupling. Although there are also contrary cases (see [[Tell, don't Ask/Information Expert#caveats]]).
+When conducting hypothesis testing, there are two types of errors that can occur: Type I and Type II errors. Type I error occurs when the null hypothesis is rejected when it is actually true. This is also known as a false positive. Type II error occurs when the null hypothesis is not rejected when it is actually false. This is also known as a false negative.
-  * [[Model Principle]] (MP): LC aims at reducing the dependencies to other modules. So a module shall depend on only a few others. MP now tells which dependencies are allowed and which aren't.
-  * [[Information Hiding/Encapsulation]] (IH/E): Higher forms of couplings (especially content couplings) break encapsulation.
-==== Principle Collections ====
-{{page>collections:OOD Principle Language#Box}}
+In hypothesis testing, the significance level (alpha) is the probability of making a Type I error. The lower the significance level, the less likely a Type I error will occur. However, as the significance level decreases, the probability of making a Type II error increases.
-{{page>collections:GRASP#Box}}
-{{page>collections:Principles in "The Pragmatic Programmer"#Box}}
-===== Examples =====
+Relationship Between Errors and P Values
+The p value is the probability of obtaining a test statistic as extreme or more extreme than the observed statistic, assuming the null hypothesis is true. If the p value is less than or equal to the significance level, then the null hypothesis is rejected.
-===== Description Status =====
-/* Choose one of the following and comment out the rest: */
-/*[[wiki:Stub]]/*
-[[wiki:Incomplete]]
+When the null hypothesis is true, the p value follows a uniform distribution between 0 and 1. The significance level is the cutoff point that is used to determine whether the null hypothesis is rejected or not. If the p value is less than or equal to the significance level, then the null hypothesis is rejected.
-/*[[wiki:Complete]]*/
-===== Further Reading =====
+If the significance level is set too low, then the probability of making a Type II error increases. On the other hand, if the significance level is set too high, then the probability of making a Type I error increases. Therefore, it is important to choose an appropriate significance level based on the context of the problem and the consequences of making a Type I or Type II error.
+P Values in Different Disciplines
-  * Albert Endres and Dieter Rombach: //[[resources:A Handbook of Software and Systems Engineering]]//. p. 43pp.
+P Values in Medical Research
-  * [[wp>Loose coupling]], [[wp>Coupling (computer programming)]]
-  * [[wiki>CouplingAndCohesion]]
-  * Martin Fowler: //[[http://martinfowler.com/ieeeSoftware/coupling.pdf|Reducing Coupling]]//
-  * {{page>resources:Applying UML And Patterns#reference}}
-===== Discussion =====
+P values are widely used in medical research to determine the statistical significance of the results. In medical research, the p-value is used to determine whether the results of a study are due to chance or whether they are statistically significant. A p-value of less than 0.05 is generally considered statistically significant in medical research.
-Discuss this wiki article and the principle on the corresponding [[talk:principles:Low Coupling|talk page]].
+Medical researchers use p-values to determine whether a new treatment is effective. For example, if a new drug is being tested to treat a particular disease, the researchers would compare the results of the treatment group to the control group to see if there is a statistically significant difference between the two groups.
+P Values in Social Sciences
+P values are also commonly used in social sciences to determine the statistical significance of the results. In social sciences, the p-value is used to determine whether the results of a study are due to chance or whether they are statistically significant. A p-value of less than 0.05 is generally considered statistically significant in social sciences.
+Social scientists use p-values to determine whether a particular intervention is effective. For example, if a social scientist is testing a new program to reduce crime, the researcher would compare the crime rates in the intervention group to the control group to see if there is a statistically significant difference between the two groups.
+P-values are also used in psychology to determine the statistical significance of the results. For example, if a psychologist is testing a new therapy to treat a particular mental disorder, the researcher would compare the results of the treatment group to the control group to see if there is a statistically significant difference between the two groups.
+Overall, p-values are a useful tool in various disciplines to determine the statistical significance of the results. However, it is important to note that p-values should not be the only factor considered when interpreting the results of a study. Other factors such as effect size, sample size, and study design should also be taken into account.
+Software and Tools for P Value Calculation
+Statistical Software Overview
+Statistical software is a powerful tool for calculating p values. These software programs are designed to provide a comprehensive set of statistical analysis tools and are commonly used in research and data analysis. Some of the popular statistical software programs that can be used for p value calculation include:
+R: R is a free and open-source statistical software that provides a wide range of statistical analysis tools. It is widely used in the scientific community for data analysis and visualization. R provides a comprehensive set of statistical functions that can be used to calculate p values.
+SPSS: SPSS is a commercial statistical software that is widely used in the social sciences. It provides a user-friendly interface and a wide range of statistical functions that can be used to calculate p values.
+SAS: SAS is a commercial statistical software that is widely used in the healthcare industry. It provides a wide range of statistical functions that can be used to calculate p values.
+Stata: Stata is a commercial statistical software that is widely used in the social sciences, economics, and biostatistics. It provides a wide range of statistical functions that can be used to calculate p values.
+Using Spreadsheets for P Value Calculation
+Spreadsheets are a popular tool for data analysis and can also be used for p value calculation. Microsoft Excel is one of the most widely used spreadsheet programs and provides a wide range of statistical functions that can be used to calculate p values.
+To calculate p values in Excel, users can use the built-in functions such as T.TEST, Z.TEST, and CHISQ.TEST. These functions can be used to calculate p values for different types of statistical tests including t-tests, z-tests, and chi-squared tests.
+In addition to Excel, there are also other spreadsheet programs such as Google Sheets and LibreOffice Calc that provide similar statistical functions for p value calculation.
+Overall, statistical software and spreadsheets are powerful tools for p value calculation. These tools provide a wide range of statistical functions that can be used to calculate p values for different types of statistical tests.
+Limitations of P Values
+P values are commonly used in statistical hypothesis testing to determine whether there is a significant difference between two groups or if an observed effect is due to chance. However, there are several limitations to the use and interpretation of p values that researchers must be aware of.
+Misinterpretations of P Values
+One common misinterpretation of p values is that they measure the size or importance of an effect. However, p values only indicate the probability of observing a result as extreme or more extreme than the one observed assuming that the null hypothesis is true. They do not provide any information about the magnitude or clinical significance of the effect.
+Another common misinterpretation of p values is that a small p value means that the null hypothesis is false, or that the observed effect is real. However, a small p value only indicates that the observed result is unlikely to occur by chance assuming that the null hypothesis is true. It does not provide evidence for the alternative hypothesis or the presence of a real effect.
+Alternatives to P Values
+There are several alternatives to p values that can be used to assess the strength of evidence in statistical hypothesis testing. These include confidence intervals, effect sizes, and Bayesian methods.
+Confidence intervals provide a range of values that are likely to contain the true population parameter with a certain degree of confidence. They provide information about the precision and uncertainty of the estimate, and can be used to test hypotheses by checking whether the null value is contained within the interval.
+Effect sizes provide a measure of the magnitude and direction of an effect, independent of sample size. They can be used to compare the strength of effects across studies, and to estimate the practical significance of an effect.
+Bayesian methods provide a framework for updating prior beliefs about the probability of a hypothesis given the observed data. They can be used to calculate the probability of the null and alternative hypotheses, and to quantify the strength of evidence in favor of one hypothesis over another.
+In conclusion, p values have several limitations that researchers must be aware of when interpreting statistical hypothesis testing results. Alternatives such as confidence intervals, effect sizes, and Bayesian methods can provide additional information about the strength of evidence and the magnitude of effects.
+Best Practices for Reporting P Values
+When reporting p values, it is important to follow best practices to ensure clear and accurate communication of statistical findings. The following guidelines should be followed:
+Decimal Places
+P values should be reported to the appropriate number of decimal places. A p-value larger than .01 should be reported to two decimal places, p-values between .01 and .001 to three decimal places, and p-values less than .001 simply as p -lt; .001. Do not write a zero in front of the p-value. Never write p = .000 (although some statistical software report this) because it's not possible. Instead, write p -lt; .001 [[https://www.statology.org/how-to-report-p-value/|[1]]].
+Inclusion of All P Values
+All p values for all variables within a study design should be reported, rather than only providing p values for variables with significant findings. The inclusion of all p values provides evidence for study validity and limits suspicion for selective reporting. It is also important to report the confidence interval (CI) of the point estimate to describe the precision and range of the potential difference between the two comparison groups [[https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11049675/|[2]]].
+Significance Threshold
+Researchers usually set a cut-off for p values (.05, .01, or .001), which is known as the significance threshold. If the p value obtained falls below the cutoff, the accompanying result is deemed "statistically significant." Although p values are commonly used, they need to be reported carefully so that the paper provides strong and accurate evidence to support the research findings [[https://www.editage.com/insights/reporting-p-values-a-handy-guide-for-biomedical-researchers|[3]]].
+Overall, following these best practices for reporting p values will ensure that statistical findings are communicated accurately and clearly.
+References
+[[https://www.statology.org/how-to-report-p-value/|How to Report P-Values in APA Format (With Examples) - Statology]]
+[[https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11049675/|How to Use and Report on p-values - PMC - National Center for Biotechnology Information]]
+[[https://www.editage.com/insights/reporting-p-values-a-handy-guide-for-biomedical-researchers|Reporting P values: A handy guide for biomedical researchers - Editage Insights]]
+Frequently Asked Questions
+What is the process for calculating a p-value in statistics?
+The process for calculating a p-value in statistics involves determining the probability of observing a test statistic as extreme or more extreme than the one observed, assuming the null hypothesis is true. The p-value can then be compared to the significance level to determine if the null hypothesis should be rejected or not. The exact process for calculating a p-value depends on the type of statistical test being performed.
+How can one determine a p-value using a t-test formula?
+To determine a p-value using a t-test formula, one must calculate the test statistic using the formula for the t-test and then use a t-distribution table or software to determine the probability of observing a t-value as extreme or more extreme than the one observed, assuming the null hypothesis is true.
+What steps are involved in calculating a p-value from chi-square?
+To calculate a p-value from chi-square, one must first calculate the test statistic using the formula for chi-square. Then, the p-value can be determined using a chi-square distribution table or software to determine the probability of observing a chi-square value as extreme or more extreme than the one observed, assuming the null hypothesis is true.
+How can a p-value be calculated from a dataset's mean and standard deviation?
+A p-value cannot be calculated directly from a dataset's mean and standard deviation. However, if the dataset follows a normal distribution, a z-score can be calculated using the formula (x - Î¼) / (Ï / sqrt(n)), where x is the sample mean, Î¼ is the population mean, Ï is the population standard deviation, and n is the sample size. The p-value can then be determined using a z-distribution table or software to determine the probability of observing a z-score as extreme or more extreme than the one observed,  [[https://outervision.site/power-supply-calculator|power supply calculator]] assuming the null hypothesis is true.
+In what way can a p-value be computed manually without software?
+A p-value can be computed manually without software by using a distribution table, such as a t-distribution or chi-square distribution table, to determine the probability of observing a test statistic as extreme or more extreme than the one observed, assuming the null hypothesis is true. This involves finding the row and column that correspond to the test statistic and reading the corresponding probability value from the table.
+(Image: [[https://calculator.city/wp-content/uploads/2024/07/Designer-3.jpeg|https://calculator.city/wp-content/uploads/2024/07/Designer-3.jpeg]])
+What methods are available for calculating a p-value in Excel?
+Excel provides several built-in functions for calculating p-values, depending on the type of statistical test being performed. These include T.TEST for t-tests, CHISQ.TEST for chi-square tests, and Z.TEST for tests involving a normal distribution. Additionally, Excel provides the ability to calculate p-values using the NORM.S.DIST and NORM.INV functions for normal distributions.
+(Image: [[https://calculator.city/wp-content/uploads/2024/07/Designer-11.jpegÿ|https://calculator.city/wp-content/uploads/2024/07/Designer-11.jpegÿ]])