First off, I recommend not using $tag as identifier when dealing with elements, since there is a difference between tag and element. The tag is the tag only, nothing else. A tag can be the start or end tag: <I_am_a_start_tag>, </I_am_an_end_tag>. An element is the two tags and everything in between them. I.e. start tag, nodevalue, end tag.
Nodevalue contains everything belonging to the element, but other elements are not included, since they are markup, not the actual data. That is, for
<outer>First part of outer's node valye <inner>inner's nodevalue, but also middle part of outer's nodevalue</inner> last part of outer's nodevalue</outer>
which means that you can not just take the p nodevalues unless you want the strong nodevalues to be included.
What you need to do is go through each p's childNodes. If it's a strong node, it's your date and you deal with that separately, if it's a text node, you concatenate it together with (possibly) other text nodes of p, but in your case you just need to take contents of the text node coming after the strong element. Do note however that this p element contains 3 child nodes: text, strong, text
<p>
<strong></strong>
text
</p>
while this contains only two: strong and text
<p><strong></strong>text</p>