Cookie检查OOP

Until yesterday I was burning my brain trying to switch from a procedural thinking to a OOP thinking; this morning I gave up. I said to my self I wasn't probably ready yet to understand it.

I started then coding in the usual way, writing a function to check if there's the cookie "logged" or not

function chkCookieLogin() {
if(isset($_COOKIE["logged"])) {
    $logged = 'true';
    $cookieValue = $_COOKIE["logged"];

    return $logged;
    return $cookieValue;
}
else {
    $logged = 'false';

    return $logged;
}
}

$result = chkCookieLogin();
if($result == 'true'){
    echo $cookieValue;
}
else {
    echo 'NO COOKIE';
}

since I run across a problem: I wanted to return two variables ($logged and $cookieValue) instead of just one. I google it and I found this answer where Jasper explains a method using an OOP point of view (or this is what I can see). That answer opened me a new vision on the OOP so I tried to rewrite what I was trying to achieve this way:

class chkCookie {
public $logged;
public $cookieValue;

public function __construct($logged, $cookieValue) {
    $this->logged = $logged;
    $this->cookieValue = $cookieValue;
}

function chkCookieLogin() {
    $out = new chkCookie();
    if(isset($_COOKIE["logged"])) {
        $out->logged = 'true';
        $out->cookieValue = $_COOKIE["logged"];

        return $out;
    }
    else {
        $out->logged = 'false';

        return $out;
    }
}
}

$vars = chkCookieLogin();
$logged = $vars->logged;
$cookieValue = $vars->cookieValue;
echo $logged; echo $cookieValue;

Obviously it didn't work at the first attempt...and neither at the second and the third. But for the first time I feel I'm at one step to "really touch" the OOP (or this is what I think!).

My questions are:

1. is this attempt correctly written from the OOP point of view?
2. If yes, what are the problems? ('cause I guess there's more than one)

Thank you so much!

Credit to @NiettheDarkAbsol for the idea of returning an Array data-type.

Using dependency injection, you can set-up an object like this:

class Factory {
    private $Data = [];
    public function set($index, $data) {
        $this->Data[$index] = $data;
    }

    public function get($index) {
        return $this->Data[$index];
    }
}

Then to use the DI module, you can set methods like so (using anonymous functions):

$f = new Factory();
$f->set('Cookies', $_SESSION);
$f->set('Check-Cookie', function() use ($f) {
    return $f->get('Cookies')['logged'] ? [true, $f->get('Cookies')['logged']] : [false, null];
});

Using error checks, we can then call the method when and as we need it:

$cookieArr = is_callable($f->get('Check-Cookie')) ? call_user_func($f->get('Check-Cookie')) : [];
echo $cookieArr[0] ? $cookieArr[1] : 'Logged is not set';

I'd also consider adding constants to your DI class, allowing more dynamic approaches rather than doing error checks each time. IE, on set() include a constant like Factory::FUNC_ARRAY so your get() method can return the closure already executed.

You can look into using ternary operators if you're confused.

See it working over at 3v4l.org.
If it means anything, here is an OOP styled approach.