Thanks John for the clarification. How about degree of polymerization? Does it influence Tm? For example, the DPs of both PE and PVC mentioned in the question I asked are the same. Can I say that having the same value of DP for both polymers cancel out the effect of higher molecular weight of PVC? According to what you stated above, can I conclude that crystallinity plays a more effective role in increasing the melting temperature of polymers after molecular weight? I mean, the chain stiffness is important if we are talking about two polymers with a similar degree of crystallinity. But, if the degree of crystallinity of the two polymers with the same DP are substantially different, the polymer with the higher crystallinity is expected to have a greater melting temperature. Am I right? 

Thanks John for your explanation. I'm a bit confused and overloaded now :( I'm not expert in polymers that is why it is hard for me to understand these basic rules. Lets ask the question in a different way. We know that there are a set of factors influencing melting temperatures of polymers: 1) Molecular weight (considering DP); 2) chain stiffness; 3) branching degree; 4) impurities; 5) density; 6) polar branches; 7) bulky branches; 8. degree of crystallinity; 9) structural regularity (isotactic, syndiotactic, etc)

So, if we consider these factors, how we can write them in the order of their effectiveness on melting temperature? Can I say the "chain stiffness" comes first, then crystallinity after structural regularity? If so, where do the other factors stand? 

Thanks for your help and time. 

Thank you very much Johan, your explanations helped me a lot

You may be raising two different questions:  the crystalline melt vs. the amorphous melt.  The discussion so far has been focused on the former.  Amorphous polymers raised to sufficiently high temperatures also pass from the rubber region to a melt or viscous flow region.  So part of the problem may be that one is not distinguishing between these two melts.  Unfortunately, "melt" is an ambiguous term.

While most of the issues raised will apply to the amorphous melt, they will be less relevant to the crystalline melt.  Molecular weight is one example:  it strongly affects the width of the rubber region and therefore the amorphous melt temperature.  The crystalline melt will be more influenced by polar interactions. 

Also keep in mind that many of these compositional issues will strongly affect the rate of crystallization that occurs, which is a potentially uncontrolled variable in your studies.  You may not be examining corresponding degrees of crystallinity, as the samples can change during storage. 

Also note that many so-called amorphous polymers are capable of some degree of crystallization -- the rate is determined by the annealing temperature the polymer is stored at.  This is a generally poorly reported phenomenon, with most writers dismissing it out of hand.   I routinely observed such crystallinity in latex copolymers of butyl acrylate, methyl methylacrylate, and methacrylic acid, as did many of my colleagues.once they started intentionally looking for it.  The crystallization rate is much slower than PE or PVC.   

So to order the influences you mentioned, you first need to separate the crystalline from the amorphous issues. 

Let's consider Tm in semicrystalline polymers. It is temperature in which molecular chains forming so far crystals losse their regular well define location and orientation in crystals and become randomly located in the amorhpous phase. Crystals form because it is energetic favourable. To melt (destroy) crystals one need to provide energy (thermal energy) to the molecule (or part of molecule) high enough to exceed energetic barrier. If so, molecule (or part of molecule in case of polymers) detaches from crystal. On one side we have temperature which is measure of energy "delivered" to polymer, on the other side we have energy of bonds between molecules (or part  in crystal. Covalent bonds between atoms in polymer chain are much stronger than crystal bonds (which are mainly van der Waals and polar in nature). Strength of crystal bonds depends on chain structure: regularity (tacticity), chain stiffness, presence of polar group etc. Macromolecules (as polymer molecules) can be attach and dettach part by part. So energy necessary to dettach such part depends on parts size. This means that crystals size play significant role (at least is well decumented in case of PE). Good prameter to characterize melting behavior of polymer is Tm0 - equlibrium melting temperature. Which is more or less  hypotetical melting temperature of inifinite perfect crystal. This temeprature is infuleced only by strength of crystal bonds.

Real Tm is lower than Tm0, because crystals are not infinite, have surface, have defects, etc.

Correlation between observed Tm and Xc, or Tm and Mw, or Tm and DP for given polymer is a effect of influence of Mw(DP) on crystallization process. Crystallization process is also influenced by temperature (isothermal crystallization) or/and time. Level of branching, presence and length of sidechains, generally speaking, molecular chain topology influences crystallization process to and crystal perfection, too.

Mw, DP, Mw/Mn, topology, crystallization temperature(time) leads to different crystals thickness and perfection and at the same time to different Xc. Crystals thickness and perfection are only factors governing Tm.

One can easily imagine samples of polymer crystallized slowly with thick crystals and sample of quenched polymer with small crystals where the same high Xc was achieved by adding nucleating agent. Both sample will have close Xc, but Tm may be quite different.

See  Hoffman J.D. works.