How to Do a PhD Like This: A Reflection

I started my PhD in September 2024. I am defending in September 2026. Exactly two years. I do not say this to boast. I say it because it taught me something about what a PhD can be when you stop following all the usual advice and start asking different questions.

Most PhD advice is not wrong. It works for many people. But it assumes you want to make a small contribution within an existing framework. I wanted something else. I wanted to build a new framework. This is what I learned along the way.

On Reading

I did not read every paper. I could not have. There are too many. Instead, I tried to read at the broadest level. I read the key papers. I read the meta-analyses. I read the reviews. I tried to understand the direction of each field — where it was going, what it assumed, what it argued about, what it left unquestioned.

Then I stopped. Not because deeper reading has no value. It has great value. But I had to make a choice about how to spend my time. I chose breadth over depth. I chose direction over mastery. I may have missed important papers. I accept that risk.

Someone will ask: “How do you know you did not miss something that would change your argument?” I do not know. No one can know. The literature is too large. The difference is that I am willing to admit this limitation openly. I read enough to build my framework. If someone shows me a paper that undermines it, I will revise. That is how scholarship should work.

On Building

I did not find a gap in the literature. I found a misdirection — or at least, I came to see the questions differently.

The existing literature on intergenerational mobility asks: “How big is the correlation between parents and children?” The implicit goal for many is to make it smaller. I began to wonder whether this was the only question worth asking. The correlation is not one thing. It is many things: genetic inheritance, financial resources, social privilege, and something else — something that comes from the daily, intimate life of families.

I came up with a new concept called Familial Relationship Goods. Shared meals. Bedtime stories. Authoritative guidance. The practices that make a family a family. I do not claim this concept is original. Others have written about family intimacy. But I tried to give it an empirical grounding it had not received before.

I operationalised the transmission effect of Familial Relationship Goods as: the environmental spillover of parents’ exposure to randomness on the child’s gene–environment system, operating through the direct, co-resident family relationship.

This sentence took me two years to think, and two months to write it down. It is dense. It is not perfect. But it is mine. I did not borrow it from anyone. I tried to build it from the ground up. I may have failed in ways I do not yet see. That is for others to judge.

On Estimating

I did not let the literature determine my conclusions. But neither did I ignore it. I read the evidence. I tried to synthesise it. I made a judgment. I estimated based on my own reading and reasoning.

The literature suggests the causal effect of parental education on child education is about 0.1 to 0.15. There is disagreement about what this means. Some say it is genetic. Some say it is environmental. Some say it is bias. I looked at the twin studies. I looked at the IV studies. I looked at the meta-analyses showing convergence. I looked at the critiques of gene-environment interplay. I built a framework about the gene-environment system and randomness, and the intergenerational transmission from that system. Accordingly, I made a correction to the current estimates: the most reasonable estimate of the permissible transmission channel is about 0.15.

I may be wrong. Future research may show 0.05 or 0.25. That would not surprise me. I am not claiming certainty. I am claiming that this is the best judgment I can make given the evidence I have. And the normative framework — that some non-zero transmission is permissible as the price of family intimacy — does not depend on the exact number. It only requires that some transmission exists.

On Time

I finished in less than two years. This was not because I worked harder than everyone else. I worked about as hard as a normal PhD student. But I worked differently, and I had help.

I started writing early. I wrote badly at first. Then I revised. I tried to keep the structure clear. I tried to keep the argument tight. I cut things that did not serve the central claim.

But this is only one side of the story. It also comes from my supervisors, who are very productive, active, and supportive. Not everyone has this privilege. I am grateful for it.

There are things I did not have time to do. There are literatures I did not engage deeply. There are nuances I cannot speak to. I tried to acknowledge these limitations honestly.

On Success

I do not know if my thesis will be considered a success. That depends on what you mean by success.

If success means a perfect thesis, I have not succeeded. My thesis has limitations.
If success means unanimous approval, I have not succeeded. Some will disagree.
If success means a job offer immediately after graduation, I do not know yet.

But if success means building something new — something that was not there before — then I have succeeded a little.
If success means finishing in a reasonable time without burning out, I have succeeded.
If success means learning to trust my own judgment while remaining open to criticism, I have succeeded enough for now.

That is all I can claim.

That is how I did this PhD. It worked for me. It may not work for you. But if you are the kind of person who wants to build something new rather than add a brick to an old wall, perhaps something here will be useful.

I wish you luck. You will need it. But you will also need judgment, honesty, and the willingness to be wrong. Those matter more.