Indexing pages that canonical urls should ignore

[onlinesource]

This problem could relate to any site, but I just happen to be operating a Magento eCommerce shopping cart.

I recently scraped my site using Screaming Frog and noticed that a lot of pages are being indexed twice, such as:

mydomain.tld/contact and mydomain.tld/contact/

Since the same canonical url of is listed mydomain.tld/contact/ is listed at both the mydomain.tld/contact/ and mydomain.tld/contact url, why is it being indexed?

I notice that not every page is indexed twice but some pages, categories and products are but I can't put my finger on why.

I have run online test and confirmed by canonical urls are working.

[lucy24]

<tangent>
Is it physically possible to reach both URLs, with and without trailing slash? There are plenty of good reasons to nip this in the bud. Choose one form-- personally I prefer the without-slash version-- and redirect all requests for the other form. Even if the "canonical" is interpreted correctly, it still means every search engine is crawling every page twice. You have to stop them from even crawling. Once they learn that every "wrong" URL will meet a speedy 301, you've effectively doubled your crawl budget.
</tangent>