A statistical speculation check determines whether or not totally different populations have the identical distribution of a categorical variable. It assesses if noticed pattern information are in step with the idea that the populations are homogeneous with respect to that variable. As an example, one would possibly use this to look at if totally different age teams have the identical preferences for various manufacturers of espresso. The null speculation assumes that the distributions are the identical throughout populations, whereas the choice speculation means that they’re totally different.
This kind of evaluation is efficacious in varied fields, offering a framework for evaluating inhabitants traits and figuring out potential variations. It aids in understanding relationships between categorical variables and group membership, which might inform decision-making. Traditionally, the event of this methodology allowed researchers to carefully check assumptions about inhabitants similarities, transferring past subjective observations. Its software spans from market analysis to social sciences, contributing to evidence-based conclusions.
Understanding the basic ideas, calculation strategies, and assumptions related to this statistical instrument is vital for correct software and interpretation. Subsequent sections will delve into the precise steps concerned in performing this check, together with information preparation, the computation of the check statistic, and the dedication of statistical significance. Sensible examples will additional illustrate its use and spotlight potential limitations.
1. Populations
The idea of “populations” is key to this statistical process, representing the distinct teams being in contrast for similarities within the distribution of a categorical variable. The check’s validity hinges on the cautious definition and collection of these populations, as any bias or non-representativeness can considerably skew the outcomes. For instance, when inspecting shopper preferences throughout totally different geographical areas (the populations), a failure to precisely signify the demographics of every area may result in inaccurate conclusions about homogeneity.
The check assesses whether or not noticed variations within the distribution of the specific variable throughout the populations are statistically vital or merely resulting from probability. With out clearly outlined populations, it turns into unimaginable to border the null and different hypotheses successfully. In a medical examine investigating the effectiveness of a brand new drug throughout totally different age teams, the age teams represent the populations. If the drug exhibits various levels of effectiveness, this check can decide if these variations are statistically vital, suggesting that age influences the drug’s efficacy.
In abstract, the correct identification and characterization of populations are important precursors to the legitimate software of this methodology. Understanding the traits and potential biases inside every inhabitants is vital for deciphering the check outcomes and drawing significant conclusions. Disregarding the nuances of inhabitants definition introduces the danger of producing flawed insights that will have sensible implications throughout varied disciplines, from healthcare to market evaluation.
2. Categorical Variable
The explicit variable varieties the cornerstone of the evaluation; it’s the attribute upon which the populations are in contrast on this check. Its presence dictates the appropriateness of the check itself, as this statistical methodology is particularly designed to evaluate distributional variations throughout teams based mostly on qualitative, slightly than quantitative, attributes. With no categorical variable, the framework for evaluating populations dissolves, rendering the check inapplicable. As an example, if the target is to find out whether or not totally different advertising and marketing campaigns yield various ranges of buyer satisfaction (categorized as ‘happy,’ ‘impartial,’ or ‘dissatisfied’), then buyer satisfaction constitutes the specific variable.
The character of the specific variable immediately influences the development of the contingency desk, which serves as the first information enter for the check. The variable’s classes outline the rows or columns of the desk, with cell entries representing the noticed frequencies of every class inside every inhabitants. If, for instance, one goals to determine whether or not political affiliation (Democrat, Republican, Unbiased) differs throughout varied age demographics, political affiliation is the specific variable, and the contingency desk would show the counts of people in every age group figuring out with every political celebration. The accuracy and completeness of the specific variable’s categorization immediately affect the validity of the check outcomes.
In conclusion, the choice and definition of the specific variable is a vital step that determines the applicability and interpretability of the check’s findings. It establishes the premise for evaluating populations and extracting significant insights. Any ambiguity or misclassification throughout the categorical variable can result in inaccurate assessments of homogeneity, highlighting the necessity for rigorous consideration to its definition and measurement. The understanding of its position is paramount for acceptable implementation and interpretation of the check.
3. Noticed frequencies
Noticed frequencies represent the empirical information collected from every inhabitants beneath investigation and are a vital enter for a chi-square check of homogeneity. These frequencies signify the precise counts of observations falling into every class of the specific variable for every inhabitants. The accuracy and reliability of those frequencies immediately affect the validity of the check; errors in information assortment or categorization can result in inaccurate conclusions in regards to the homogeneity of populations. For instance, when learning shopper preferences for various manufacturers of smartphones throughout varied age teams, the noticed frequencies can be the variety of people in every age group preferring every model. These frequencies type the premise for comparability between the age teams. An inaccurate depend of preferences, both resulting from sampling bias or information entry errors, would immediately affect the check outcomes, probably resulting in incorrect conclusions about whether or not smartphone preferences differ throughout age teams.
The chi-square check makes use of noticed frequencies to calculate anticipated frequencies beneath the null speculation that the populations are homogeneous. The check statistic then quantifies the discrepancy between the noticed and anticipated frequencies, primarily assessing whether or not the deviations are bigger than what can be anticipated by probability alone. If the noticed frequencies deviate considerably from the anticipated frequencies, the check supplies proof in opposition to the null speculation, suggesting that the populations are usually not homogeneous with respect to the specific variable. As an example, in a scientific trial evaluating the effectiveness of two totally different therapies for a illness, the noticed frequencies can be the variety of sufferers in every remedy group who expertise a constructive final result, a adverse final result, or no change. If the noticed frequencies of constructive outcomes are considerably larger in a single remedy group in comparison with the opposite, this means a statistically vital distinction in effectiveness between the 2 therapies.
In abstract, noticed frequencies are an indispensable part of the chi-square check of homogeneity. They signify the empirical basis upon which the check is constructed, and their correct assortment and illustration are essential for acquiring legitimate and significant outcomes. The check’s capability to detect variations between populations hinges on the precision of those noticed frequencies, making cautious consideration to information assortment and administration paramount. Understanding the position and significance of noticed frequencies is important for successfully making use of and deciphering the chi-square check of homogeneity in varied analysis contexts.
4. Anticipated frequencies
Throughout the framework of a check of homogeneity, the idea of “anticipated frequencies” is central to assessing whether or not noticed information deviate considerably from what can be anticipated if the populations beneath examination have been certainly homogeneous with respect to the specific variable of curiosity.
-
Calculation Underneath the Null Speculation
Anticipated frequencies are computed based mostly on the idea that the distributions of the specific variable are similar throughout all populations. These frequencies are derived by multiplying the row and column totals of the contingency desk and dividing by the general desk whole. This calculation supplies the frequency one would anticipate to watch in every cell if there have been no precise affiliation between inhabitants membership and the distribution of the specific variable. As an example, if inspecting buyer satisfaction ranges (excessive, medium, low) throughout totally different retailer areas, the anticipated frequency for ‘excessive satisfaction’ at a selected location is calculated as (whole ‘excessive satisfaction’ responses throughout all areas * whole responses from that location) / (whole responses general).
-
Comparability with Noticed Frequencies
The check statistic quantifies the mixture distinction between the noticed and anticipated frequencies. A bigger discrepancy suggests stronger proof in opposition to the null speculation of homogeneity. The check measures whether or not the noticed deviations from what is predicted by probability alone are substantial sufficient to warrant rejecting the idea that the populations have the identical underlying distribution. For instance, if a considerably higher-than-expected variety of clients at one location report ‘excessive satisfaction,’ this disparity contributes to a bigger check statistic, probably indicating a distinction in satisfaction ranges throughout areas.
-
Affect on the Check Statistic
The magnitude of the anticipated frequencies immediately influences the check statistic. Cells with small anticipated frequencies can disproportionately affect the chi-square worth, probably resulting in inflated check statistics and inaccurate rejection of the null speculation. To mitigate this challenge, a typical rule of thumb is that not more than 20% of cells ought to have anticipated frequencies lower than 5, and no cell ought to have an anticipated frequency lower than 1. If these circumstances are usually not met, different strategies, corresponding to combining classes or utilizing Fisher’s actual check, could also be extra acceptable. Contemplate a situation the place a really uncommon class is included within the evaluation; even a small distinction between noticed and anticipated frequencies in that class can have a big impact on the check statistic.
-
Position in Figuring out Statistical Significance
The calculated check statistic, together with the levels of freedom (decided by the variety of classes and populations), is used to find out a p-value. The p-value represents the likelihood of observing a check statistic as excessive as, or extra excessive than, the one calculated, assuming the null speculation is true. A small p-value (usually lower than 0.05) supplies proof to reject the null speculation and conclude that the populations are usually not homogeneous with respect to the specific variable. Due to this fact, anticipated frequencies play an oblique, but essential, position on this dedication, as they’re important to calculating the check statistic that finally results in the p-value.
In conclusion, anticipated frequencies are usually not merely computational artifacts however slightly integral elements of the check, offering a baseline in opposition to which to judge noticed information. Their correct calculation and correct interpretation are important for drawing legitimate inferences in regards to the homogeneity of populations.
5. Contingency desk
The contingency desk serves as the basic information construction for conducting a check of homogeneity. It organizes categorical information, tabulating the frequencies of observations throughout totally different populations and classes of a specified variable. With no correctly constructed contingency desk, the mandatory calculations for this statistical check are usually not doable. The desk’s rows usually signify the populations being in contrast, whereas the columns signify the totally different classes of the specific variable. Every cell throughout the desk incorporates the noticed frequency, which is the depend of observations belonging to a specific inhabitants and falling into a selected class. For instance, a examine evaluating buyer satisfaction (happy, impartial, dissatisfied) throughout totally different retailer areas would use a contingency desk to show the variety of clients in every location falling into every satisfaction class. The check then evaluates whether or not the distribution of satisfaction ranges is constant throughout all retailer areas.
The construction of the contingency desk immediately impacts the calculation of anticipated frequencies, an important step within the check. Anticipated frequencies are derived beneath the null speculation, assuming that the populations are homogeneous with respect to the specific variable. The check compares these anticipated frequencies with the noticed frequencies within the contingency desk to calculate a check statistic. A big distinction between noticed and anticipated frequencies, as indicated by a big check statistic, supplies proof in opposition to the null speculation, suggesting that the populations are usually not homogeneous. As an example, if a contingency desk reveals a disproportionately excessive variety of happy clients at one retailer location in comparison with what can be anticipated beneath homogeneity, this might contribute to a bigger check statistic and a decrease p-value, probably resulting in the conclusion that buyer satisfaction varies considerably throughout areas.
In abstract, the contingency desk is an indispensable instrument for conducting a check of homogeneity. It supplies a structured framework for organizing and analyzing categorical information, enabling the computation of anticipated frequencies and the evaluation of statistical significance. The accuracy and readability of the contingency desk immediately affect the validity and interpretability of the check outcomes. A well-constructed desk ensures that the check is appropriately utilized and that the conclusions drawn are supported by the information. Challenges can come up from small pattern sizes or sparse information throughout the desk, probably resulting in unreliable outcomes. Addressing these challenges via acceptable information assortment and evaluation strategies is essential for the efficient use of the check.
6. Levels of freedom
Levels of freedom are a vital parameter within the check, influencing the interpretation of the check statistic and the dedication of statistical significance. The check, designed to evaluate the homogeneity of categorical variable distributions throughout totally different populations, depends on levels of freedom to correctly contextualize the calculated check statistic. This parameter displays the variety of unbiased items of knowledge obtainable to estimate a inhabitants parameter. Within the context of this specific check, levels of freedom are decided by the variety of classes within the categorical variable and the variety of populations being in contrast. Particularly, levels of freedom equal (variety of rows – 1) (variety of columns – 1), the place rows signify the populations and columns signify the classes of the variable. For instance, if evaluating political affiliation (Democrat, Republican, Unbiased) throughout 4 totally different age teams, the levels of freedom can be (4-1) (3-1) = 6. This worth signifies the form of the chi-square distribution used to judge the importance of the check statistic. A misunderstanding of levels of freedom can result in incorrect conclusions in regards to the homogeneity of the populations.
The calculated check statistic, which quantifies the distinction between noticed and anticipated frequencies, should be interpreted in relation to the levels of freedom. A bigger check statistic is extra prone to be statistically vital when the levels of freedom are decrease, because the vital worth for rejection of the null speculation decreases. Conversely, for a similar check statistic, a better diploma of freedom will elevate the vital worth and should result in a failure to reject the null speculation, regardless of obvious variations within the noticed information. Suppose two separate research are performed to evaluate whether or not there are variations in product choice between women and men. The primary examine examines solely two product choices, whereas the second examine examines 5. Even when each research yield the identical check statistic, the second examine, with its larger levels of freedom, would require a bigger check statistic to achieve statistical significance.
In conclusion, levels of freedom are usually not merely a computational part of the check, however a basic aspect in figuring out statistical significance. An understanding of this parameter and its affect on the chi-square distribution is important for drawing legitimate inferences in regards to the homogeneity of populations. A failure to account for levels of freedom may end up in misinterpretation of the check statistic and inaccurate conclusions relating to the similarity of distributions. Due to this fact, correct calculation and conscious consideration of levels of freedom are important for the suitable software and interpretation of this statistical methodology.
7. Check statistic
The check statistic is a central aspect within the analysis of the null speculation. Within the context of a check of homogeneity, it serves as a quantitative measure of the discrepancy between noticed information and the information anticipated if the populations being in contrast have been actually homogeneous.
-
Calculation Methodology
The check statistic in a check of homogeneity is calculated utilizing a formulation that sums the squared variations between noticed and anticipated frequencies, every divided by the corresponding anticipated frequency. This calculation produces a single numerical worth representing the general divergence of the pattern information from the null speculation. As an example, if inspecting buyer preferences for 3 totally different manufacturers throughout two demographic teams, the check statistic would combination the variations between noticed preferences and the preferences anticipated if each demographic teams had the identical model preferences.
-
Interpretation and Distribution
The check statistic follows a chi-square distribution beneath the null speculation, with levels of freedom decided by the variety of classes and populations being in contrast. A bigger worth of the check statistic signifies a higher departure from the null speculation. The calculated worth is then in comparison with a vital worth from the chi-square distribution or used to find out a p-value, offering proof to both reject or fail to reject the null speculation. For instance, a excessive check statistic, relative to the levels of freedom, means that the noticed variations in buyer preferences throughout the 2 demographic teams are unlikely to have occurred by probability alone, thus questioning the idea of homogeneity.
-
Components Influencing the Check Statistic
A number of elements can affect the magnitude of the check statistic, together with pattern dimension, the variety of classes within the categorical variable, and the magnitude of variations between noticed and anticipated frequencies. Bigger pattern sizes have a tendency to extend the check statistic, as do bigger variations between noticed and anticipated values. Small anticipated frequencies in some cells may also disproportionately inflate the check statistic. These concerns are vital when deciphering the outcomes and figuring out whether or not the noticed variations are virtually vital along with being statistically vital. As an example, even a small distinction in model choice may yield a excessive check statistic if the pattern dimension is sufficiently massive, necessitating a cautious analysis of the sensible implications of the findings.
-
Limitations and Assumptions
The validity of the check statistic depends on sure assumptions, together with the independence of observations and sufficiently massive anticipated frequencies in every cell of the contingency desk. Violations of those assumptions can compromise the accuracy of the p-value and result in inaccurate conclusions. For instance, if the anticipated frequency in any cell is lower than 5, the check statistic could not precisely comply with a chi-square distribution, and different strategies, corresponding to Fisher’s actual check, could be extra acceptable. Moreover, the check statistic solely supplies proof of affiliation, not causation. Even when a big distinction is discovered, it doesn’t essentially indicate that one inhabitants immediately influences the distribution of the specific variable in one other.
The check statistic is thus an integral part of a check of homogeneity, offering a quantifiable measure of the variations between populations. Its calculation, interpretation, and the evaluation of its validity require a cautious consideration of pattern traits, distributional assumptions, and potential limitations, all of which contribute to the right software and understanding of the check.
8. P-value
The p-value is a basic part within the decision-making course of inside a check of homogeneity, representing the likelihood of observing pattern information as excessive as, or extra excessive than, what was really noticed, assuming the null speculation is true. The null speculation, on this context, asserts that the populations being in contrast have the identical distribution of the specific variable. A small p-value supplies proof in opposition to this assertion, suggesting that the noticed variations in distributions are unlikely to have arisen by probability alone. The calculation of the p-value is immediately linked to the check statistic calculated from the noticed and anticipated frequencies throughout the contingency desk. This hyperlink is the inspiration for assessing statistical significance. For instance, in market analysis evaluating model preferences throughout totally different age teams, a small p-value derived from a check signifies that the noticed variations in model preferences between the age teams are statistically vital, implying that age influences model choice.
The sensible significance of the p-value lies in its capability to tell choices in varied fields. It permits researchers and analysts to quantify the power of proof in opposition to the null speculation, enabling them to make knowledgeable judgments about whether or not to reject the idea of homogeneity. The traditional threshold for statistical significance is a p-value of 0.05, that means that there’s a 5% probability of observing the information if the null speculation is true. In medical analysis, this threshold could be used to find out whether or not a brand new remedy has a considerably totally different impact in comparison with a management remedy throughout totally different demographic teams. A p-value lower than 0.05 would recommend that the remedy impact shouldn’t be uniform throughout all demographic teams. Nevertheless, it is vital to notice {that a} statistically vital p-value doesn’t routinely indicate sensible significance. The magnitude of the impact and the context of the analysis are equally vital concerns.
Deciphering the p-value requires cautious consideration of the examine design, pattern dimension, and potential confounding elements. A low p-value doesn’t show that the null speculation is fake, however slightly means that the noticed information present ample proof to reject it. Conversely, a excessive p-value doesn’t show that the null speculation is true, however slightly suggests that there’s inadequate proof to reject it. Challenges in deciphering the p-value can come up from small pattern sizes, which can lack the facility to detect true variations between populations. Moreover, relying solely on the p-value with out contemplating the impact dimension and confidence intervals can result in deceptive conclusions. Finally, the p-value serves as an important instrument within the check, however its interpretation requires a nuanced understanding of its limitations and context.
Incessantly Requested Questions
This part addresses widespread inquiries relating to the statistical methodology for figuring out whether or not totally different populations have the identical distribution of a categorical variable. These questions and solutions intention to make clear its purposes, limitations, and interpretations.
Query 1: What distinguishes the chi-square check of homogeneity from the chi-square check of independence?
The check of homogeneity examines whether or not a number of populations share the identical distribution of a categorical variable. The check of independence, conversely, assesses whether or not two categorical variables are related inside a single inhabitants. Knowledge assortment strategies additional distinguish these assessments. The check of homogeneity entails choosing samples from a number of populations, whereas the check of independence entails a single pattern the place two variables are measured for every topic.
Query 2: What are the important thing assumptions underlying the check?
This check depends on a number of assumptions. First, the information should be randomly sampled from the populations of curiosity. Second, the observations should be unbiased of each other. Third, the anticipated frequencies in every cell of the contingency desk needs to be sufficiently massive. A typical rule of thumb is that not more than 20% of cells ought to have anticipated frequencies lower than 5, and no cell ought to have an anticipated frequency lower than 1. Violations of those assumptions can compromise the validity of the check outcomes.
Query 3: How are levels of freedom calculated for this check?
Levels of freedom are calculated as (variety of rows – 1) multiplied by (variety of columns – 1), the place rows signify the populations being in contrast and columns signify the classes of the specific variable. This worth determines the form of the chi-square distribution used to evaluate the statistical significance of the check statistic.
Query 4: What does a statistically vital consequence indicate?
A statistically vital consequence (usually a p-value lower than 0.05) suggests that there’s ample proof to reject the null speculation of homogeneity. This suggests that the populations being in contrast would not have the identical distribution of the specific variable. Nevertheless, statistical significance doesn’t essentially equate to sensible significance. The magnitude of the impact and the context of the analysis must also be thought of.
Query 5: What are some widespread purposes of this check?
This check finds software throughout numerous fields. In market analysis, it could examine buyer preferences throughout totally different demographic teams. In healthcare, it could assess whether or not the distribution of illness incidence differs throughout varied geographical areas. In social sciences, it could study whether or not attitudes towards a specific challenge differ throughout totally different political affiliations. These are however a number of examples demonstrating the breadth of its applicability.
Query 6: What limitations needs to be thought of when deciphering the outcomes?
A number of limitations needs to be thought of. The check solely assesses whether or not populations have totally different distributions; it doesn’t clarify why these variations exist. Moreover, it’s delicate to pattern dimension, with bigger samples probably resulting in statistically vital outcomes even for small variations. Moreover, the check assumes that the information are categorical; it’s not acceptable for steady variables. Lastly, the check supplies proof of affiliation, not causation.
The check is a precious instrument for evaluating the distributions of categorical variables throughout totally different populations. Nevertheless, its acceptable software and interpretation require cautious consideration of its assumptions, limitations, and the precise context of the analysis query.
The following part will present an in depth, step-by-step information on the best way to carry out a check of homogeneity, together with information preparation, calculation of the check statistic, and interpretation of the outcomes.
Sensible Ideas for Making use of the Check
This part provides steering on maximizing the utility and accuracy when making use of the statistical methodology for evaluating the distribution of a categorical variable throughout a number of populations. Adherence to those pointers enhances the reliability of findings and facilitates sound conclusions.
Tip 1: Guarantee Sufficient Pattern Dimension: Inadequate pattern sizes can compromise the facility of the check, probably resulting in a failure to detect true variations between populations. An influence evaluation, performed previous to information assortment, can decide the mandatory pattern dimension to realize a desired degree of statistical energy. As an example, when evaluating shopper preferences throughout totally different areas, be certain that the pattern from every area is massive sufficient to signify the inhabitants precisely.
Tip 2: Confirm Independence of Observations: This statistical methodology assumes that observations are unbiased. Violation of this assumption can result in inflated check statistics and inaccurate conclusions. In research involving paired or clustered information, different analytical strategies that account for dependence needs to be thought of. Contemplate a scenario the place information is collected from members of the identical family; the responses are prone to be correlated and violate this independence assumption.
Tip 3: Deal with Low Anticipated Frequencies: Low anticipated frequencies in some cells of the contingency desk can distort the check statistic and result in inaccurate p-values. If greater than 20% of cells have anticipated frequencies lower than 5, or any cell has an anticipated frequency lower than 1, contemplate combining classes or utilizing different statistical strategies, corresponding to Fisher’s actual check. For instance, if learning the connection between occupation and political affiliation, and one occupation class has only a few respondents, contemplate merging it with an analogous class.
Tip 4: Clearly Outline Categorical Variables: Ambiguous or poorly outlined classes can introduce bias and compromise the interpretability of the outcomes. Be sure that the classes are mutually unique and collectively exhaustive, and that the standards for assigning observations to every class are clearly specified. If analyzing buyer satisfaction, be certain that the classes (e.g., “happy,” “impartial,” “dissatisfied”) are well-defined and unambiguous.
Tip 5: Interpret Statistical Significance in Context: A statistically vital consequence doesn’t routinely indicate sensible significance. Contemplate the magnitude of the impact, the cost-benefit ratio of any potential interventions, and the potential for confounding elements. In healthcare analysis, a brand new remedy could present a statistically vital enchancment in comparison with a management, but when the advance is minimal and the associated fee is excessive, it is probably not virtually vital.
Tip 6: Report Confidence Intervals and Impact Sizes: Along with the p-value, report confidence intervals and impact sizes to offer a extra full image of the magnitude and precision of the noticed results. Impact sizes, corresponding to Cramer’s V or Phi coefficient, quantify the power of the affiliation between populations and the specific variable. These measures present precious context past the binary determination of statistical significance.
Tip 7: Consider Potential Confounding Variables: Confounding variables can distort the connection between populations and the specific variable. Contemplate together with potential confounders as management variables within the evaluation or utilizing stratification strategies to account for his or her affect. For instance, if inspecting the connection between schooling degree and revenue, contemplate controlling for age, as age can affect each schooling degree and revenue.
The following pointers, when rigorously thought of and applied, improve the validity and interpretability of findings. Such cautious software contributes to extra knowledgeable choices and a extra nuanced understanding of the relationships between populations and categorical variables.
Having coated these concerns, the dialogue transitions to the ultimate part, summarizing the core ideas and highlighting the overarching significance of the tactic.
Conclusion
This exposition has detailed the “chi-square check of homogeneity,” a statistical methodology for assessing whether or not totally different populations share a typical distribution of a categorical variable. The dialogue encompassed its underlying assumptions, calculation procedures, the interpretation of its outcomes, and potential limitations. Key concerns highlighted embody the significance of sufficient pattern dimension, the independence of observations, and the correct calculation of anticipated frequencies. Understanding the nuances of this statistical instrument is essential for researchers and analysts throughout varied disciplines.
Rigorous software and conscious interpretation of the outcomes are important for drawing legitimate inferences and informing sound choices. The “chi-square check of homogeneity” stays a precious instrument in comparative evaluation, and ongoing consciousness of its capabilities and constraints will promote extra even handed and evidence-based conclusions throughout numerous fields of inquiry. Continued refinement of knowledge assortment strategies and analytical strategies will additional improve the robustness and applicability of this methodology sooner or later.
