Correlation Directional Index Calculator

Introduction

The Correlation Directional Index (CDI) calculator evaluates the directional alignment and co-movement between two numerical datasets. This tool is essential for researchers analysing whether variables $X$ and $Y$ tend to deviate from their respective means in the same or opposite directions across $n$ observations. It provides a non-parametric perspective on relationship strength through directional agreement and binomial sign testing.

What this calculator does

This tool processes two paired numerical sequences to determine the frequency of co-movement relative to the dataset means. Users input raw values for both variables, and the calculator identifies valid pairs where non-zero deviations occur. The output includes the Correlation Directional Index, a count of co-positive and co-negative matches, and a binomial sign test $p$ -value to determine if the observed directional relationship is statistically significant.

Formula used

The calculation first determines the mean of each dataset and evaluates the deviation of each pair. A match is recorded if the product of deviations is positive. The primary index is the ratio of matches to valid pairs. Statistical significance is assessed using a binomial cumulative distribution function where the null hypothesis assumes a probability of $0.5$ for directional agreement.

CDI = \frac{n_{matches}}{n_{valid}}

P (X \geq k) = \sum_{i = k}^{n} (\binom{n}{i}) {0.5}^{n}

How to use this calculator

1. Enter the numerical values for the $X$ variable into the first input field.
2. Enter the corresponding numerical values for the $Y$ variable into the second field.
3. Select the desired number of decimal places for the resulting metrics.
4. Execute the calculation to view the index, sign test results, and trend visualisation.

Example calculation

Scenario: A researcher is studying environmental measurements to determine if fluctuations in soil temperature and moisture levels exhibit consistent directional alignment over several days of observation.

Inputs: $X$ values: $10, 20, 30$ ; $Y$ values: $5, 15, 25$ .

Working:

Step 1: $μ_{x} = \frac{10 + 20 + 30}{3} = 20; μ_{y} = \frac{5 + 15 + 25}{3} = 15$

Step 2: $Deviations : (- 10, - 10), (0, 0), (10, 10)$

Step 3: $Valid Pairs = 2; Matches = 2$

Step 4: $CDI = \frac{2}{2} = 1.00$

Result: 1.00

Interpretation: The result indicates a perfect positive directional agreement between the datasets.

Summary: All valid pairs moved in the same direction relative to their respective means.

Understanding the result

A CDI value near $1.0$ suggests strong positive directional agreement, whereas a value near $0.0$ indicates inverse alignment. The $p$ -value assesses whether this alignment is likely due to chance; values below $0.05$ suggest that the directional relationship is statistically significant and not merely a random occurrence.

Assumptions and limitations

The method assumes the data consists of paired observations from two continuous or discrete variables. It does not require normal distribution, but observations must be independent. Pairs with zero deviation from the mean are excluded from the index calculation, which may impact results in very small datasets.

Common mistakes to avoid

A common error is confusing the CDI with Pearson's correlation coefficient; the CDI only measures directional co-movement, not the magnitude of change. Another mistake is ignoring the $p$ -value, as a high index in a small sample may not be statistically significant. Ensure that both $X$ and $Y$ datasets contain an equal number of values.

Sensitivity and robustness

The calculation is highly sensitive to the mean, as individual data points are categorised based on their position relative to the average. Outliers can shift the mean and alter the directional status of multiple pairs. However, because it uses binary directional classification, the index is robust against extreme values that do not cross the mean threshold.

Troubleshooting

If an error appears regarding dataset mismatch, verify that the count of $X$ values exactly matches the count of $Y$ values. If the CDI is $0$ or results are excluded, check if all data points are identical to the mean, as the tool requires variance to determine direction.

Frequently asked questions

What does "Excluded (No deviation)" mean?

This occurs when a data point is exactly equal to the mean of its dataset, making it impossible to determine a positive or negative direction for that specific pair.

How is the p-value calculated?

The tool uses a Binomial Sign Test, calculating the probability of obtaining the observed number of matches (or more) if the true probability of a match was exactly 50%.

Can this handle negative numbers?

Yes, the calculator processes any real numerical values between -1e12 and 1e12, focusing on their directional movement relative to the calculated mean.

Where this calculation is used

This statistical measure is used in academic fields such as social research and population studies to analyse co-movement between trends. For example, researchers may use it in sports analysis to see if player performance metrics and team victory margins move in the same direction over a season. In environmental science, it helps model how different atmospheric variables correlate directionally. Unlike standard correlation, it provides a simplified view of alignment that is useful for introductory data modelling and non-parametric hypothesis testing where directional consistency is the primary interest.

Results are based on standard mathematical and statistical methods and may involve rounding or approximation. If precise accuracy is required, please verify results independently. See full disclaimer.