Study Skills and Exam Strategies

​HOW TO DRINK FROM A FIRE-HOSE WITHOUT DROWNING

Successful study strategies in medical and pharmacy school
April Apperson, UC San Diego SOM

Why should I change my study strategies?

The practice of medicine and pharmacy requires physicians and pharmacists to rapidly and accurately access a lot of information in their heads that is originally acquired from a great variety of sources.

Only change your strategies if you're NOT satisfied with your exam scores.

Material presented in medical or pharmacy school is not conceptually more difficult than many rigorous undergraduate courses, but the volume flow rate of information per hour and per day is much greater – it has frequently been described as "drinking from a fire-hose."

  • Everyone admitted to any medical or pharmacy school has developed study strategies successful for an undergraduate informational flow rate
  • Unfortunately, those developed by most undergraduate students are not efficient enough for the fire hose flow rate. (Virtually no one develops efficient study habits until their performance requires it.)

Efficient studying – the best use of your limited time – requires active learning, not passive.

If you are comfortable while you are studying, it almost certainly isn't active.

Active studying requires not knowing (uncomfortable), then working it out.

Passive studying is comfortable - just not efficient.

  • Active learning requires constantly making decisions about the material – "Is this important?" "How is this different from the other examples given for the same process?" "Where does this fit into the ‘big picture'?" "What is the exact definition of this term?"
  • Passive reading of pages of text or "going over" notes (even with a highlighter) and "absorbing" the information is inefficient – if you have enough time, it will eventually work, but it usually isn't adequate for the fire hose.

Changing a habit that has worked for many years isn't easy, so don't be surprised if you need to hear the same things many times – and don't be afraid to ask to hear it again!

What are The fundamental PROCESSES IN  successful active studyING?

  1. Identify the important information – answering the eternal question of "what's important here?"
  2. Organize the information – create chunks of information in patterns that facilitate memorization and emphasize the connections needed for application questions, e.g., for differential diagnosis. 
  3. Memorize the information – this is an active process and not just reading it over enough times. Don't put this off because "I'll just forget it again". Memorize your chunks and patterns as soon as possible, then review the main points each time the topic appears again. It's annoying, but more efficient, and the information will get into long-term memory more effectively.
  4. Apply the information to more complex situations – practice this before the exam.

WhAT ARE THE GOALS OF AN ACTIVE STUDY PLAN?

  1. CREATE as quickly as possible the study sheets, aka study aids, that together contain the needed information organized into chunks and patterns that promote memorization and integration.
  2. MEMORIZE and SYNTHESIZE that information by applying it in all available practice formats, e.g., practice questions, small group conference or problems sessions, quiz questions, etc.

What are the SUGGESTED steps of this plan?

  1. Before lecture, devote ≈ 10 minutes per lecture hour to identify and memorize the "big picture". These steps focus on lectures presented from powerpoint slides, but can be adapted for texts.
    • The goal of pre-lecture work is to have a basic roadmap of the lecture in your head – if you get lost during a complex part of lecture, you won't be lost for the rest of the lecture.
      • Use the outline to identify the key headings - then memorize the number of headings and then the actual headings (knowing the number helps!) to create your basic roadmap.
      • Use the remaining time to understand the basic logic behind the sequence of headings - this will also help you remember them.
  2. During lecture, take notes that emphasize connections and what the lecturer says is important
    • If possible, take handwritten notes electronically (e.g., tablet) or on a slide printout - handwriting activates analytical centers that typing does not.
    • If practical, annotate power point slides or lecturer notes, rather than taking completely new notes; use abbreviations and symbols to increase what you can take down.
      • Add the "take-home message" for each graph or chart; clearly label the axes.
      • Note lecturer comments that identify or integrate key points, especially in the intro or summaries.
      • Always note information that helps you decide between options, e.g., circumstances that indicate when one reflex or response will outweigh another, or in what circumstances one DNA test is used rather than another, etc.
    •  Let unimportant information go! Don't rewrite data already present; use a single slash across material the lecturer says is not key (you can still read it); leave out "color commentary" or "cool new stuff" that is outside the scope of the class.
  3. After lecture, on the same day (if feasible), edit your notes and begin organization.
    • Your lecture notes become the messy but complete back-up document – DON'T rewrite them.
    • Edit notes in a single pass by using the text (if needed) and learning objectives (if useful) to:
      • Identify/add any missing major headings or subheadings that reinforce your "road map".
      • Circle specific terms and definitions in a different color – these will be used both for reference and for keying memorization of the material.
      • Decide and identify what is needed for your study sheets and what can be left out.
        • Scan learning objectives (if specific enough to be useful) to help these decisions.
        • Don't get caught up working through non-relevant information on figures in slides taken from an article or text that include parts not needed for the key point.
        • If a key point or definition doesn't make sense, use the text or online references - but stay focused!
    • If a lecturer discusses something you remember from an earlier lecture, check it out! If it includes a different perspective or additional information, add the relevant information so your study aid will have all of it in one place.
    • For topics that lend themselves to reasonably simple organizations, e.g., flow diagrams, sketch out the organization.
  4. Within a few days, find time to complete your lists, charts, flow diagrams and pictures – these study aids form the "final draft" of your information that you will use to study for the exam (don't go back to your original notes!).
    • Memorize the organizations actively as you complete each one, and review them frequently to consolidate them in your memory (see Tips below).
    • Apply the information in practice and quiz questions well before the exam. Use them to work out a method for approaching complex problems that might appear on the exam.

TIPS ON ORGANIZING, MEMORIZING AND APPLYING THE MATERIAL

General tips:

  • Design your study aids to illustrate a logical sequence within the material and emphasize connections or comparisons.
    • The analysis of the information needed for this is active studying! The resulting study aids work with known aspects of memory, which decreases the burden of "rote" memorizing.
    • People remember stories rather than unrelated facts. The order of headings in a chart or flow diagram, or the placement of the elements within a diagram should reflect a time-line or location or relationships in a logical order - this can act as the "story".
    • Studies show that the more connections within the information being organized or studied, the better and longer you will remember it. And virtually all people remember visual patterns better than paragraphs of text.
    • Flow diagrams or pictures or graphs are more effective that writing out a sequence of sentences or long descriptions.
    • For detailed material, start with a simpler, more memorizable overview that identifies each topic by a key word or phrase to establish the connections and the "story". Create further flow diagrams or organizations for each topic within the overview that needs more detail.
  • For lists, charts/tables and flow diagrams, use the "rule of 5": no more than five items under one heading. If you need more, then decide how the items are logically divided into subgroups and create subheadings.
  • Don't recreate the wheel. If you find a good chart in some text or other source, add it to your summaries.  Customize it by crossing out non-core material and adding any needed material from lecture — use color to make changes stand out. 
What organization should I use?
Some material lends itself to certain organizations, but much is just personal preference.
  • Deciding on the headings and subheadings in a list or table, or how the information is organized in a flow diagram, or where it is located in a picture or diagram, is the ACTIVE part of the process.
    • Sketch out different possibilities without adding the actual content, using the rule of 5 (or 6).
    • Try for one that flows logically and visually emphasizes distinguishing features between different examples or processes.
  • Different organizations have different strengths.
    • Tables clearly compare/contrast related processes or structures, but they invite simple transcription (passive!) and can be text-heavy and therefore harder to memorize.
    • Flow diagrams are very useful for organizing related topics and subtopics that have examples, but they can get too complex and hard to follow if the relationships aren't clear.
    • Graphs are very useful to present complex relationships in a simple visual pattern, but memorize the labels on the axes and not just the line shapes!
  • Finding connections between different topics aids memorization and application
    • Use equivalent headings or orders of headings for separate tables or flow diagrams on related topics – it both aids memorization and emphasizes connections within the material that help with application.
      • You can create and use a "default" heading sequence for some information:
      • For neurotransmitters, hormones, etc.: structure, synthesis, receptor and signaling, biological effect and regulation
      • For molecular processes: fundamentals, initiation, elongation, termination, processing.
    • Create "big picture" charts, flow diagrams or pictures that organize KEY related topics from multiple lectures – the topic charts from each lecture will provide the detail.
    • Don't hesitate to include key information on different summaries, especially if they are organizing the material from different points of view or at different levels of detail.
Memorize actively and use frequent short reviews.
  • Organized material is useless unless it is accessible within your memory - preferably long term.
    • Getting information into long-term memory requires multiple repetitions and active memorization – "going over" or "reading over" material is NOT efficient.
    • Cramming doesn't work in the long term! Courses build on each other and you need information from earlier courses to understand and apply material from later courses.
    • The most efficient memorization methods are also the LEAST painful.
  • Active memorization builds in steps: how many, then key words/phrases, then chunks.
    1. Memorize how many items (e.g., headings) there are so you know if you have them all.
    2. Memorize the headings in order - use biological logic, visualization, and/or mnemonics - before starting on the content.
    3. For the content, assign a key word or short phrase for each needed chunk of information, e.g., a cell in a table, or a piece of information in a flow diagram.
      • Memorize these for one row, column, or region, using the heading sequence or connections to reinforce the material.
      • Close your eyes and recite that piece - how many, headings, key words - until its clear.
    4. Add any needed further information to each key word/phrase and memorize that combination, and repeat two processes under step three.
    5. Continue until you think you have the full study aid memorized.
  • Self-test each full organization on paper or a whiteboard:
    • Cover the original and write it out on a blank paper/board (don't be pretty, but don't cheat).
    • Throw away or erase what you have just done (this is the tough part – get rid of it) and look at the original.
      • If you are confident that you wrote out all parts correctly, great1 (DON'T COMPARE)
      • If there is a part you're not sure about, DON'T COMPARE to what you wrote. Rememorize it and self-test again.
    • This method emphasizes what you don't know, while comparing old and new will just emphasize what you do know (more reassuring, but less effective!)
  • Actively review the key word/phrase version of a topic organization at every opportunity.
    • Every time an earlier topic or concept is mentioned in a lecture or practice question, close your eyes and quickly review the topic organization - how many, headings and key words/phrases.
    • Mentally review graphs or pictures or diagrams at other times, e.g., waiting for laundry to dry.
    • This sets the information in your memory and builds practice in the mental sequences you'll use answering questions under time pressure
  • When studying with others, quizzing each other can be helpful (and fun), but beware that you might be using subliminal cues to help answer questions without truly mastering the material. When studying along, explaining material out loud to yourself can also be helpful, but beware of the very common tendency to verbally "hand-wave" around areas you aren't actually clear on.  Always check yourself with the active review described above.

Practice application early - do NOT wait until you believe you can get all the practice Qs right!

  • Practice questions are diagnostic tools best used to master the material BEFORE you know it.
    • Use the stems AND the answer choices (correct and incorrect) to help identify the level of detail you need to know, and how much integration and/or application is expected.
    • If you got a question CORRECT, identify exactly why each incorrect choice was wrong and why it was included as a potential choice for that question. Actively review your study organizations that include each incorrect choice. Predict a question stem that uses the incorrect choice as its focus.
    • If you got a question wrong, actively memorize the study aid that includes the question topic (and not just the part that applies directly to the correct answer).
      • Work out an approach or method for that type of question – "where do I start"?
      • If the material includes clinical cases or scenarios, predict application questions by asking: Given data (clinical or molecular), determine: i) the most likely cause of the symptoms or test results, ii) the best assay/technique to use, iii) the mutation that could cause the results, or iv) the calculation needed to determine the risk of having the disease, or to generate the pressure or flow asked about, etc.
  • Ungraded weekly quizzes are useful for gauging mastery of the material, but don't put them off beyond the correct weekend because it encourages falling even further behind; analyze the questions as described above to help mast the material.
  • Use practice or quiz questions more than once.
    • Retaking earlier quiz questions after a couple of weeks can point out content that was only in short-term memory. Actively rememorize it!
    • Even if you remember the correct answer, work through the approach or method for the question and use it to mentally review the relevant information. Use this to decrease the time needed to answer complex questions.
    • Change values or other features in the question to create a new version and work through it.
  • Where do I find time for all this?

    1. Successful high-volume studying relies on good investment strategies:
      1. Finding the "big picture" before lecture is easily put off, but it pays off by increasing learning during lecture and facilitating design of organizations after lecture.
      2. Focusing on key points and NOT getting caught up in unneeded detail on slides or in "color commentary" slides will provide more time.
      3. Creating study aids takes time, but you'll learn the material with connections and insight that will let you answer those application questions; study aids also pay off with more time during the inevitable finals crunch because you won't be leafing through endless pages of slides wondering where that information was (and what they all mean).
      4. Memorizing as you go and frequent review of earlier concepts as you study pays off immediately because you'll need less time to study for subsequent lectures. You'll be able to recognize connections and correlations that lead to more effective organizational aids, and you'll remember the material both more effectively and longer.
    2. There is more time available in a day than you think – don't let "studying all that" help you procrastinate.
      1. Divide your studying into a series of short tasks so you don't feel so overwhelmed, and so you don't feel the need to wait until you have 2 or 3 hours to study. Convincing yourself to start on one chart or diagram is much easier than finding a large block of time. And once you're started, it's easier to keep going.
      2. Use small bits of time while your clothes are drying or 1/2 of your lunch hour, or while the rice is cooking for dinner for active studying tasks.
      3. Use all the "extra" time you can in the first week to be caught up in lectures, because the courses (blocks) are short and there is much to cover.
      4. Be VERY careful about "robbing Peter to pay Paul". It's tempting to skip lecture or to put off viewing a recorded lecture to finish an upcoming project - but try to keep it to a minimum. These are typically bad investments because the lectures build on each other, so you'll probably need even more time to master the material

What are the most common problems MS1 and P2 students have ON exams

  1. Clarity and precision in definitions and concepts vs. approximate definitions derived from context.
    • Often, students generate their own general concepts or definitions from context (after all, that's how we learn to speak) – but this doesn't provide enough clarity to analyze and correctly answer the questions.
    • Medical terminology and equations are very precise – being "close enough" usually isn't sufficient.
  2. Familiarity with material vs. mastery of the material.
    • "Familiarity" refers to recognizing the logic provided by someone else – as when leaving a good lecture, you can say, "yeah, that made sense."
    • Mastery of the material requires integration and memorization of sufficient detail that the information can be successfully applied to a new situation.

Frequently Asked Questions/Frequently Heard Comments (FAQ):

TOPIC:  "How do I know what will be on the exam and practice for it?"

"How do I know what will be on the exam?" or "How do I know how much detail to learn?"

  1. How do I know how much detail to learn from lectures?
    1. If the learning objectives (LOs) are "task-oriented" and specific, use them to check your study organizations - DON'T write them out (this takes too much time and is all text!).
      • While generating study aids, use a quick review of the LOs for the relevant topic(s) to help decide what to include or what is unneeded detail.
      • When study aids for a major topic or a lecture are done, review the LOs and make sure you could answer each from your organizations.
      • Task-oriented learning objectives focus on the level of material likely to be assessed, which for most courses is the lecture and for a few may include a required text. So using outside Web sources to "answer" learning objectives is not efficient and often encourages students to go well beyond the lecture's intent.
    2. Check your lecture notes – they should emphasize the lecturer's context, which includes any comments by the lecturer on what is important.
    3. If the learning objectives are unfocused, the lecture outline, introductory and summary slides, and key point slides can help provide needed "big picture" concepts. Identify and include examples or conditions that support and illustrate these. If the lecturer has provided any additional tables or charts, use them!!
    4. Other indications that information is important:
      1. If the concept is mentioned in more than one lecture, it's probably important. The same topic may be addressed from different points of view in different lectures, but the exam question on the topic may integrate all those points of view.
      2. If a clinical example or disease is described or explained (not just referred to in passing), you may well need to know the molecular or biochemical or genetic abnormality and the biochemical or physiological consequences of the abnormality.
      3. If the instructor gives a sample calculation, take note. If an accompanying table has more examples, fill it in.
      4. If the course includes conferences with assigned questions, be sure to analyze and answer all the questions that other students presented. There is a lot of difference between tracking logic outlined by the lecturer or text, and applying principles in exam questions.
  2. Use any practice or sample problems to identify the level of "necessary detail" (see the question below); use quiz questions in the same way, and then expand that level to apply to any topics or calculations in your study aids that were not asked about on the quiz.
  3. Assume that all lectures – including introductory lectures – will be asked about on the exam. In most cases, the questions on an exam are balanced to be approximately proportional to lecture/conference hours per topic. Of course, it doesn't always work out that way, but it is a place to start.

"But I need to save the quizzes or practice questions to test myself after I know all the material."

Don't save practice questions to test yourself just before the exam – it is then too late to modify your studying!! You may do well and be complacent, or do poorly and be a nervous wreck – neither helps!

Practice exams or questions are a vital piece of the answer to the questions, "How much of detail do I need?" and "What kind of questions will they ask?"

Start using practice questions as you create the summary for a particular topic — after setting the headings in your summary, check over the practice questions on that topic to make sure you haven't missed anything.

You can use practice questions again – see the answer to the question below.

"How do I use quiz questions to help my studying?"

  1. Don't just count up how many you got right – analyze each correct AND incorrect question thoroughly.
    • Make sure you understand why EACH possible answer is right or wrong and how each wrong answer could be made correct and what topic it was referring to (why did they include it?).
    • If you didn't get it right, go back and memorize the entire chart or summary that contained it.
  2. Ask yourself if there are any other examples that could easily be used with the same format. Locate the information for each quiz question in your study aids – if it isn't there, add it and any equivalent information to any related categories.
  3. Try writing a few possible questions yourself, then answering them (or trade with a friend).  This is a very powerful technique because you have to analyze the material, know the big picture and know details.

"I need more practice questions to study from; where can I get them?"

Used properly, relatively few quiz and sample questions are plenty for preparation — see question above.

"I did well on the weekly quiz, but then I didn't do well on this exam."

The exam will contain different question and covers more material to integrate and apply.
  • Did you take the earlier quizzes again before the exam? This can identify material that needs to be rememorized or integrated with later material.
  • Did you take the ungraded quizzes "open book"? If so, the predictive value of their scores is low.
  • Did you allow yourself as much time as you needed on the quiz? This can allow you to develop useful methods for approaching difficult questions, but it doesn't provide the time-pressure you may feel during the exam.
Thorough analysis of the quiz questions (see above), taking the quizzes a second time with time pressure and after an extra week or more to "forget" material, and predicting a few harder questions will all help with your preparation for the exam.

TOPIC: "I didn't do well on the exam – I need help with my test-taking strategies."

"I understood the material, but I had trouble answering the questions — my problem is test-taking."
Test-taking strategies can always be improved and can help each student display what he or she really does know, but usually most of the problem with less than stellar exam scores is passive studying strategies. Two common types of problems are listed below.
  1. Clarity of definitions or concepts vs. those derived from context.
    • Students often generate their own general concepts or definitions from context (that is how we learn to speak) – but this doesn't provide enough clarity to analyze and correctly answer the questions.
    • Medical terminology and equations are very precise – being "close enough" often isn't sufficient. A "general idea" of a definition often leads to choosing a similar but incorrect answer.
    • Knowing the exact definitions and equations very well also increases exam speed, allowing more time for analytical questions.
  2. Familiarity with material vs. mastery of the material.
    • "Familiarity" refers to recognizing the logic provided by someone else – as when leaving a good lecture, you can say, "yeah, that made sense."
    • Mastery of the material requires memorization of integrated and organized material with sufficient detail so that needed information can be successfully applied to new situation.
    • Knowing how much detail to learn is difficult, and varies with each class – see the first FAQ above.
"I have trouble with those trick questions."
Many "trick" questions aren't tricks at all — they just require careful reading (so will prescriptions, etc.) and care in answering. In general:
  • Focus on the data in question by highlighting (exam software) relevant information in the stem.
  • Jot down key information on the scratch paper in a quick organization if needed and do any needed calculations.
  • Have a general idea of the possible answers before you read the answers, then find the most correct answer among them.
  • Try to have a "back-up" alternative logic to verify your answer, other than that in the answer itself; emphasizing organization and connections during studying helps this technique this a lot.
Common specific problems that increase the chance of missing a question include:
  • Choosing the first familiar answer just to "move on" in the exam.
    • Rely on clarity, not familiarity, and write out the appropriate equation or relationship.
    • Use the time generated by clarity on straightforward questions to be methodical on all questions.
  • Choosing a familiar answer because you don't recognize the specific scenario or condition described in the stem.
    • The question is probably testing fundamental principles in a new application. Think basics!
    • What situation that you've heard of might be analogous to this condition? What fundamental principles or equations apply in this situation?
    • Choose the answer that clearly works with those principles or equations - not just one the "makes sense".
  • Choosing the first correct statement, even though it is not the best answer to the question.
  • Be sure the statement answers the stem and is true under the conditions listed.
  • Be sure you have not added a condition or assumption to the stem that isn't actually there.
  • Assuming an "easy" answer is too easy and choosing a different answer.
    • A well-written exam includes questions with varying levels of difficulty, which often include testing basics. If you have generated organized study aids and memorized them, some questions will probably be "obvious" to you (this isn't a trick).
    • Write out the appropriate equation or relationship to confirm the "easy" answer.
    • In this situation, students frequently impose conditions or assumptions that are NOT present in the stem to legitimize an incorrect answer - beware of this.
Unfocused questions ("Which of the following is true concerning _____?) require a different technique:
  • Read each answer as a continuous statement with the stem or as an answer to the stem and decide if it is true or false.
  • Use the highlighter provided in the exam software to mark off incorrect answers and decide which of the remaining is most accurate.
"Except" or "which of the following is false" questions also require a different technique:
  • These require extra care in answering because it is very easy to distracted by your analysis of the stem and just pick the next correct statement (which is NOT the correct choice for the question).
  • If the exam software allows, mark every answer as T or F; if the exam software doesn't allow this, write T or F (not the answer choices) in the correct order on your scratch paper.
  • If you see multiple Ts marked as you look at the answers, you're more likely to remember you need to choose a false "F" statement and the answer to the question.

"I always get it down to 2 answers and the always choose the wrong one."
This is a very real problem for many students – it's not just a misperception. At least two different processes seem to contribute to consistently doing this.
  1. Students unconsciously equate effort with accuracy. So if it takes 3 minutes to justify one answer and only 30 seconds to justify another, the 3 minute answer is often chosen – after all, that effort must mean something (or you're just tired of trying to figure it out, so you pick the one you just spent a lot of effort working on). Of course this usually means that it took more assumptions and tortuous reasoning to justify the 3-minute answer, which makes it less likely to be true.
  2. Students also frequently give more validity to impressions of "what your body wants to do" or "what the cell wants to do" than to actual known relationships or equations that describe reality. Often, a student will say something like, "Well, I knew that stroke volume times heart rate equals cardiac output, and answer 'b' fit that equation, but I thought that stroke volume should be really important in determining myocardial oxygen consumption (MV02) because the heart wants to move all that blood, so I chose answer 'c'." Notice the difference between "knew" and "should". 
When deciding between two potential answer choices, try staying with the answer that you first chose as correct (using clarity rather than familiarity) unless you recognize a serious flaw in your logic or suddenly remember a fact or equation!
If this continues to be a problem, especially if exam question choices are unfocused (see earlier topic), it may be worth considering each answer with a 5-part scoring system to help analyze the question.
  • Read the question stem once carefully and then begin reading each answer carefully. As you read each answer choice the first time, write its score (described below) in order on scratch paper.
    1. TT = definitely true – you are confident of the fact or can write out the equation or can visualize the information from your notes or summaries that support it.
    2. T = probably true – you aren't confident, but something tells you it's probably true.
    3. ? = no clue – don't be afraid to use this if you don't know!
    4. F = probably false – you aren't confident, but something tells you it's probably false.
    5. FF = definitely false – you are confident of the fact or can write out the equation or can visualize the information from your notes or summaries that support the fact this is false.
    • It's important to assign a score on your first impression, since this is usually based on fundamentals and often correct.
    • The more you continue to analyze a question when you are uncertain, the more likely you are to tangle up your analysis and end up choosing an incorrect answer.
  • Scan the answer scores for their pattern.
    1. If only one is TT or T and the others are ?, F or FF – no problem, pick the correct answer.
    2. If there are both TT and T or both T and ? (or both F and FF on an "EXCEPT" question), re-read the stem and the each answer as a single unit. Try to utilize biological or physiological mechanisms or equations in your analysis. Don't change a T into a TT, or a "?" into a T, or an F into an FF unless you can write the supporting evidence on your exam page. That rule will allow you to change answers when you need to and minimize answer changing for the wrong reasons.
When I change answers, I always seem to change them from right answers to wrong answers.
If you don't have a good reason to change an answer, the first answer will most likely be correct. Your first answer choice is usually based (often subconsciously) on fundamentals – that's good. If you start over-thinking a question, you're likely to mislead yourself by getting sidetracked into unlikely logic paths with more assumptions and tortuous reasoning. These are the same ones that take up a lot of time and make the answer less likely to be true. So:
  1. If you can write out the reason to change your answer, do so. This should take care of those times that you really recognized an error in your logic or in an equation, or a later question gave you some information you had forgotten.
  2. If you can't write out the reason, don't change the answer. This should minimize the problems explained in the first paragraph.
You may lose a question occasionally using this advice, but overall, you'll end up with better scores. (Several studies have concluded that changing answers on multiple choice answer exams is beneficial, but if you read the studies, these studies don't discriminate between changing answers for valid reasons versus not. This advice takes care of both issues.)